Fluorogenic reporter enables identification of compounds that inhibit SARS-CoV-2

Yang, Junjiao; Xiao, Yinghong; Lidsky, Peter V.; Wu, Chien-Ting; Bonser, Luke R.; Peng, Shiming; Garcia-Knight, Miguel A.; Tassetto, Michel; Chung, Chan-I; Li, Xiaoquan; Nakayama, Tsuguhisa; Lee, Ivan T.; Nayak, Jayakar V.; Ghias, Khadija; Hargett, Kirsten L.; Shoichet, Brian K.; Erle, David J.; Jackson, Peter K.; Andino, Raul; Shu, Xiaokun

doi:10.1038/s41564-022-01288-5

Download PDF

Article
Published: 05 January 2023

Fluorogenic reporter enables identification of compounds that inhibit SARS-CoV-2

Nature Microbiology volume 8, pages 121–134 (2023)Cite this article

7248 Accesses
6 Citations
30 Altmetric
Metrics details

Subjects

Abstract

The coronavirus SARS-CoV-2 causes the severe disease COVID-19. SARS-CoV-2 infection is initiated by interaction of the viral spike protein and host receptor angiotensin-converting enzyme 2 (ACE2). We report an improved bright and reversible fluorogenic reporter, named SURF (split UnaG-based reversible and fluorogenic protein–protein interaction reporter), that we apply to monitor real-time interactions between spike and ACE2 in living cells. SURF has a large dynamic range with a dark-to-bright fluorescence signal that requires no exogenous cofactors. Utilizing this reporter, we carried out a high-throughput screening of small-molecule libraries. We identified three natural compounds that block replication of SARS-CoV-2 in both Vero cells and human primary nasal and bronchial epithelial cells. Cell biological and biochemical experiments validated all three compounds and showed that they block the early stages of viral infection. Two of the inhibitors, bruceine A and gamabufotalin, were also found to block replication of the Delta and Omicron variants of SARS-CoV-2. Both bruceine A and gamabufotalin exhibited potent antiviral activity in K18-hACE2 and wild-type C57BL6/J mice, as evidenced by reduced viral titres in the lung and brain, and protection from alveolar and peribronchial inflammation in the lung, thereby limiting disease progression. We propose that our fluorescent assay can be applied to identify antiviral compounds with potential as therapeutic treatment for COVID-19 and other respiratory diseases.

A VSV-based assay quantifies coronavirus Mpro/3CLpro/Nsp5 main protease activity and chemical inhibition

Article Open access 27 April 2022

A SARS-CoV-2 protein interaction map reveals targets for drug repurposing

Article 30 April 2020

Boceprevir, GC-376, and calpain inhibitors II, XII inhibit SARS-CoV-2 viral replication by targeting the viral main protease

Article 15 June 2020

Main

The worldwide pandemic of the respiratory disease COVID-19, caused by the coronavirus SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), has produced huge health challenges and economic loss. While effective vaccines and therapeutic monoclonal antibodies are available, a large part of the global population remains unvaccinated and without access to monoclonal therapies due to their cost and limited supply^1,2. Moreover, immune escape mutations in the spike (S) protein may weaken the efficacy of current interventions³. Thus, small-molecule drugs are urgently needed to curb progression of the COVID-19 pandemic.

The first step in the entry of SARS-CoV-2 into human cells is viral binding to the host cell, which is achieved by protein–protein interaction (PPI) between the virus surface-anchored S protein and the host cell receptor angiotensin-converting enzyme 2 (ACE2)^4,5,6,7. The trimeric S protein is cleaved into S1 and S2 subunits; subunit S1 contains a receptor-binding domain (RBD) that binds the extracellular peptidase domain of ACE2. The coronavirus enters the host cells^8,9 and viral genomic RNA is released into the cytoplasm. After viral RNA replication, mature virions are assembled and released via budding from host cells.

Inhibition of the PPI between the S protein and ACE2 may therefore block the entry of SARS-CoV-2 into human cells, prevent infection and ultimately remedy COVID-19 disease. Indeed, this PPI is a major target for therapeutic intervention and researchers have developed antibodies, de novo designed proteins, peptides and engineered ACE2 to bind the S protein and inhibit viral binding and infection^10,11,12,13. Here, we report the identification of small-molecule inhibitors, including natural compounds, that block the interaction of S with ACE2, inhibit binding of SARS-CoV-2 to cells and potently suppress replication of coronavirus in host cells.

We identified the inhibitors using a reversible and fluorogenic PPI reporter that enables direct visualization of spike and ACE2 interactions in living cells. This PPI reporter has a large dynamic range, with a dark-to-bright fluorescence signal that requires no exogenous cofactors. It visualizes the interaction between S protein and its receptor ACE2 and is generalizable to imaging the interaction of spike with other receptors such as neuropillin-1. Two of the natural products identified using this method potently inhibit replication of SARS-CoV-2 in vivo and could be developed into antiviral therapeutics for treatment of COVID-19.

Results

Engineering reporters for imaging of spike–ACE2 interactions

To design a fluorogenic reporter for detecting interaction between the S protein and ACE2, we improved a green fluorescent PPI reporter that we previously engineered from UnaG^14,15. We named this PPI reporter SURF (split UnaG-based reversible and fluorogenic PPI reporter). Like its parent protein UnaG, SURF incorporates endogenous bilirubin as the chromophore and thus its genetically encoded fluorescence does not require an exogenous cofactor. To visualize interaction of two proteins, we fuse each fragment of SURF (carboxyl-terminal fragment, cSURF; amino-terminal fragment, nSURF) to the proteins of interest (Fig. 1a). When interactions of the two proteins bring the two fragments into close proximity, the two parts of SURF reconstitute and become fluorescent (Fig. 1a and Supplementary Fig. 1). Because SURF is reversible, inhibition of PPI leads to dissociation of the two fragments of SURF, resulting in the absence of fluorescence.

**Fig. 1: Engineering fluorogenic reporters for imaging of spike–ACE2 interaction in cells.**

Next, we applied SURF to the design of PPI reporters of S–ACE2 interaction by structure-based protein design and engineering. Crystal structures of complexes of the S protein, ACE2 and B⁰AT1, revealed that the PPI of S and ACE2 is mediated by the receptor-binding domain of the S protein (S^RBD)^4,5,6,7. Thus, we fused C- and N-terminal components of SURF—that is, cSURF and nSURF—to S^RBD and ACE2, respectively (Fig. 1b). The interaction of S^RBD and ACE2 brought the two fragments of SURF into close proximity so that SURF reconstituted and became fluorescent. Disruption of S^RBD–ACE2 interaction led to dissociation of the two SURF fragments, resulting in the loss of fluorescence. To design the two parts of the SURF reporter into one construct, we utilized the ‘self-cleaving’ 2 A peptide¹⁶ (Fig. 1c, upper left). We expressed this initial SURF reporter in human embryonic kidney 293 T (HEK293T) cells (Fig. 1c, lower left), which showed green fluorescence as expected although fluorescence was dim.

To make the reporter more suitable for high-throughput screening (HTS) we used structure-guided protein engineering, which increased the level of brightness by ~270-fold (Fig. 1c, right, 1d). Specifically, we engineered ACE2 by truncating the C-terminal fragment of ACE2 and found that removal of residues 616–727 improved brightness 2.2-fold (Fig. 1d). We then shortened the N-terminal fragment of ACE2 by removal of residues 1–16, which further increased brightness 3.5-fold. The truncated ACE2, including residues 17–615, was named ΔACE2. Truncation of S^RBD did not improve brightness. We next incorporated a red fluorescent protein mCherry^17,18,19 into our reporter to normalize the green fluorescence of SURF. We placed mCherry between S^RBD and cSURF, which surprisingly improved SURF brightness 30-fold (Fig. 1d). mCherry probably functions as a spacer between S^RBD and cSURF that may enhance SURF reconstitution following association of S^RBD and ΔACE2. A second mCherry between nSURF and ΔACE2 further improved brightness, leading to a SURF reporter 270-fold brighter than the initial reporter. We named this variant of SURF reporter S^RBD::ΔACE2 SURF (for simplicity, the two incorporated mCherry moieties are not shown).

Next, we examined reversibility of the S^RBD::ΔACE2 SURF system. S^RBD or ΔACE2 (each fused to near-infrared fluorescence protein IFP2)^20,21 was co-expressed to compete against S^RBD-cSURF or ΔACE2-nSURF, respectively, which would inhibit SURF reconstitution. SURF fluorescence was abolished, whereas mCherry fluorescence was stable (Fig. 1e and Supplementary Fig. 2). As a control, IFP2 expression did not perturb SURF fluorescence. Quantification of the results and calculation of a z-factor²²

$$Z = 1 - \frac{{(3\sigma ^ + + 3\sigma ^ - )}}{{\left| {\mu ^ + - \mu ^ - } \right|}},$$

gave us ~0.7 with IFP2 and IFP2 fusion to S^RBD (or ΔACE2) as positive (+) and negative (–) controls of the S and ACE2 PPI, respectively (here, σ is standard deviation, μ is mean). This indicates that S^RBD::ΔACE2 SURF is a highly robust and versatile assay for HTS of PPI inhibitors that block S–ACE2 interaction in HEK293T cells. Thus, our SURF-based approach enabled us to conduct HTS in living cells at biosafety level 2.

Screening PPI inhibitors of spike and ACE2 using SURF

Using the S^RBD::ΔACE2 SURF assay, we screened libraries of 1,516 US Food and Drug Administration (FDA)-approved drugs and 2,592 natural compounds (Fig. 2a and Extended Data Fig. 1a). First, the reporter was transfected into HEK293T cells followed by addition of the compounds with incubation for 24 h. Next, we imaged the SURF and mCherry fluorescence of the cells. The degree of PPI inhibition was calculated based on SURF fluorescence normalized by mCherry. From the FDA-approved drug library, 15 drugs showed >40% PPI inhibition with P < 0.01 (Extended Data Fig. 1b); for the natural compound library, 52 compounds showed >50% PPI inhibition with P < 0.01 (Fig. 2b).

**Fig. 2: HPS of natural compounds that inhibit spike and ACE2 interaction.**

To confirm the inhibitory activity of the FDA-approved drugs, an independent imaging of 15 drugs was conducted under the same conditions. We selected the top ten drugs (Extended Data Fig. 1c) and characterized their inhibition kinetics by time-lapse imaging using SURF. Seven drugs that inhibited the PPI within 6 h (Extended Data Fig. 1d) were selected for antiviral activity measurement (see below). Finally, the drugs showed little effect on a SURF control where cSURF and nSURF were directly linked, suggesting that they have little effect on the reporter itself (Supplementary Fig. 3).

For the identified natural compounds, their activity was confirmed by a second round of imaging (Supplementary Fig. 4). Next, we narrowed the number of hits by shortening incubation time to 6 h: 25 compounds exhibited >50% PPI inhibition (Fig. 2c) while the other 27 either displayed little inhibition (19 compounds) or strong cytotoxicity based on morphological changes (Supplementary Fig. 5). We next characterized inhibition kinetics: 19 compounds inhibited the PPI within 2 h (Fig. 2d). Among these, bruceine D, bruceine A, homoharrington, anisomycin and gamabufotalin blocked the PPI with τ ~0.5–0.7 h. The remaining compounds showed slower inhibition (Supplementary Fig. 6). Finally, the top 19 compounds had little effect on the SURF control (Supplementary Fig. 7).

Antiviral activity against SARS-CoV-2 in cells

We next selected the top-performing drugs and natural compounds against the PPI and determined their antiviral activity using SARS-CoV-2 in Vero-E6 cells (Fig. 3a). Six of these drugs inhibited SARS-CoV-2 at 5 μM (Fig. 3b), with the natural compounds inhibiting SARS-CoV-2 replication at 1 or 2 μM (Fig. 3c). Strikingly, gamabufotalin, anisomycin and bruceine A exhibited >6 log reduction in viral titre at 1 µM while at 0.1 μM concentration they inhibited SARS-CoV-2 by 10–1,000-fold (Fig. 3d). Interestingly, bruceine D, a compound structurally related to bruceine A, exhibited no inhibition at 0.1 μM (Supplementary Fig. 8).

**Fig. 3: Antiviral activities of the identified natural compounds and drugs against SARS-CoV-2.**

Next, we determined dose–response curves (Fig. 3e). Gamabufotalin showed the most potent antiviral activity, with half-maximal effective concentration (EC₅₀) ~0.003 μM, followed by anisomycin (0.016 μM) and bruceine A (0.054 μM). Interestingly, bruceine A showed fourfold higher potency than bruceine D. These compounds were largely not cytotoxic in Vero cells at EC₅₀, with selectivity index (50% cytotoxicity concentration (CC₅₀)/EC₅₀) >10. Other natural compounds and FDA-approved drugs showed less potent antiviral activity, with EC₅₀ at submicromolar (Fig. 3e) or micromolar range (Supplementary Fig. 9). The top three inhibitors were further confirmed using antibodies against the nucleocapsid (N) or spike. At 200 nM, gamabufotalin, anisomycin and bruceine A potently blocked SARS-CoV-2 (Fig. 3f,g); these compounds also blocked SARS-CoV-2 variants Delta and Omicron (Extended Data Fig. 2).

Antiviral activity during SARS-CoV-2 infection

To further understand mode of action, we examined the effects of the natural compounds at three time points before or after SARS-CoV-2 infection (Fig. 4a). For the first condition we preincubated Vero cells with inhibitors for 3 h, followed by the addition of SARS-CoV-2 (multiplicity of infection (MOI) = 5) for adsorption at 4 °C for 1 h. We then removed the solution and added fresh medium containing the compounds (Fig. 4a (i), −4 h). For the other two conditions, compounds were added at 1- or 4-h post-infection (h.p.i.) (Fig. 4a (ii), +1 h and (iii), +4 h samples). Cells were then incubated at 37 °C; 16 h later both supernatant and cells were collected for plaque assay and quantitative PCR with reverse transcription (RT–qPCR).

**Fig. 4: Identified natural compounds inhibit early stages of SARS-CoV-2 infection.**

Plaque assay-based measurement of viral particles in the supernatant showed that anisomycin and bruceine A reduced viral titre by between four and five logs and gamabufotalin by between two and three (Fig. 4b (i), −4 h). For the +1 h samples, anisomycin and bruceine A reduced viral titre by four logs and gamabufotalin by between one and two. The inhibitory effect was further reduced in the +4 h regimen. Anisomycin and bruceine A showed viral titre reduction by two logs and gamabufotalin by less than one.

Consistent with virus production measured by plaque assay, viral genome copy number measured by RT–qPCR in infected cells showed that, for −4 h samples, anisomycin, bruceine A and gamabufotalin reduced vRNA load by between five and six, between four and five and between two and three logs, respectively. Inhibitory activity was reduced in +1 h samples and further reduced in +4 h samples (Fig. 4c). These results indicate that inhibitors act on the early stages of the viral infection.

To test whether these natural compounds inhibit cellular binding of SARS-CoV-2, we conducted an immunofluorescence-based binding assay. We preincubated Vero cells with each inhibitor (2 μM) for 3 h, followed by the addition of SARS-CoV-2 (MOI = 100) for adsorption at 4 °C for 1 h (Fig. 4d). We then removed the solution, washed the cells and paraformaldehyde (PFA) fixed and stained the samples with antibodies against the N protein. Immunofluorescence of the N protein was decreased by around tenfold, suggesting that SARS-CoV-2 viral particle binding to cells is largely reduced in the presence of inhibitors (Fig. 4d). These results indicate that inhibitors act by blockage of SARS-CoV-2 binding to host cells.

To further examine whether inhibitors block the PPI of S and ACE2, we utilized the replication-competent recombinant vesicular stomatitis virus (rVSV) encoding the spike in place of the original G glycoprotein (rVSV-enhanced green fluorescent protein (eGFP)-SARS-CoV-2)²³. We examined the effects of the inhibitors on pseudovirus in ACE2-expressing A549 cells (Fig. 4e). Bruceine A, anisomycin and gamabufotalin showed strong inhibition, with an EC₅₀ of 0.018, 0.018 and 0.005 μM, respectively. Finally, we showed that our compounds had no effect on a control virus, a Sindbis virus (SINV; Supplementary Fig. 10). The above results suggest that these compounds exert their activity by blockage of the PPI of S and ACE2.

Anti-SARS-CoV-2 activity in primary human cells

Next, we determined the antiviral potential of the natural compounds gamabufotalin, bruceine A and anisomycin in human primary cells relevant to SARS-CoV-2 infection in humans²⁴. We used primary human nasal epithelial cells (HNEpCs) and primary human bronchial epithelial cells (HBEpCs), both of which were cultured in an air–liquid interface. We preincubated HNEpCs with 2 μM natural compounds for 3 h before infection (Fig. 5a). Next, we infected HNE cells with SARS-CoV-2 for 1 h at 37 °C and incubation for 48 h (both stages in the presence of the 2 μM inhibitors). Following viral infection, cells were fixed and stained with anti-N and anti-S antibodies. Immunofluorescence imaging showed that the natural compounds largely reduced SARS-CoV-2 infection in HNEpCs (Fig. 5b,d). We then conducted similar experiments in HBEpCs with the compounds at 15, 50 and 200 nM and 1 μM (Fig. 5c and Supplementary Fig. 11a). Quantification of HBEpC imaging data showed that all three compounds reduced infection even at 15 nM (Fig. 5e and Supplementary Fig. 11b). We also confirmed the inhibitory effect of gamabufotalin and anisomycin in HNE cells using plaque assays and RT–qPCR (Extended Data Fig. 3). These inhibitors therefore have potent antiviral activity against SARS-CoV-2 infection in primary human airway epithelial cells.

**Fig. 5: Inhibition of SARS-CoV-2 in primary HNEpCs and HBEpCs by the identified natural compounds.**

We also examined the effects of the natural compounds when given before or after infection of SARS-CoV-2 in HBEpCs (Extended Data Fig. 4). First we preincubated HBEpCs with inhibitors for 3 h, followed by the addition of SARS-CoV-2 for adsorption for 1 h. We then removed the solution and added fresh medium containing the compounds (Extended Data Fig. 4a (i), −4 h). For the other two conditions, compounds were added at 1 or 4 h.p.i. (Extended Data Fig. 4a (ii), +1 h and (iii) +4 h samples). Cells were incubated at 37 °C for 72 h and fixed for immunofluorescence staining against SARS-CoV-2 N and S proteins. Our data showed that, for all three compounds, −4 h samples showed complete inhibition of SARS-CoV-2 infection, +1 h samples showed ~90% inhibition but +4 h samples showed no significant inhibition (Extended Data Fig. 4b,c). These results are consistent with those for Vero cells. Our data indicate that these inhibitors act on the early stages of viral infection in HBEpCs. We also examined the toxicity of the natural compounds in HBEpCs. Following incubation with gamabufotalin, bruceine A or anisomycin at 2 μM for 72 h, no cytotoxicity was observed in HBEpCs (Supplementary Fig. 12).

Anti-SARS-CoV-2 activity in mouse models of infection

We next determined antiviral activity of the potent inhibitors gamabufotalin and bruceine A in mouse models of SARS-CoV-2 infection. K18-hACE2 mice²⁵ were infected intranasally with SARS-CoV-2 (10³ plaque-forming units (PFU) per mouse) in the absence (vehicle control, solvent) or presence of inhibitor. Treatment was repeated on days 1 and 2 (Fig. 6a). We directly examined the effects of bruceine A and gamabufotalin on virus replication. At day 3 post-infection, lungs and brain were harvested for analysis. Both compounds reduced viral titre by between two and three logs in the lung (Fig. 6b). Bruceine A treatment reduced SARS-CoV-2 titre to undetectable levels in brain tissues (<10 PFU g^–1; n = 6 animals). Similarly, we were unable to detect virus in the brains of four out of six animals treated with gamabufotalin and we observed a reduction of between one and two logs in viral titre for the remaining two animals. Immunofluorescence imaging analysis revealed similar antiviral activity for bruceine A and gamabufotalin in lung sections. Immunofluorescence of anti-N and anti-S indicated that the lungs of both bruceine A- and gamabufotalin-treated animals showed markedly fewer N- and S-positive cells than vehicle controls (Fig. 6c,d).

**Fig. 6: Inhibition of SARS-CoV-2 infection in mouse models by bruceine A and gamabufotalin.**

We next examined histopathological lung changes by staining sections with haematoxylin and eosin (H&E). In vehicle-treated samples we detected both alveolar and peribronchial inflammation. Bruceine A-treated lung showed little or no inflammation in alveolar and peribronchial areas, similar to that from animals not infected with SARS-CoV-2 (Extended Data Figs. 5 and 6). Gamabufotalin-treated lung also revealed little alveolar inflammation and largely reduced peribronchial inflammation.

We also examined whether inhibitors could block coronavirus after viral infection in mice. We first infected wild-type C57BL6 mice with another strain of SARS-CoV-2 (B.1.1.7) via intranasal inoculation without co-application of inhibitor. Treatment with bruceine A and gamabufotalin were then given twice, at 24 and 48 h.p.i., via intranasal inoculation-elicited protection (Extended Data Fig. 7a). Lungs were harvested for analysis at 72 h.p.i. Bruceine A and gamabufotalin reduced lung virus titre by 0.8 and one log, respectively (Extended Data Fig. 7b). Immunofluorescence of anti-N and anti-S showed that both drugs largely reduced N- and S-positive cells in lung (Extended Data Fig. 7c,d).

Histopathological changes in the C57BL6/B.1.1.7 model were mild, but perivascular and alveolar inflammation was still detected in the vehicle-treated group (Extended Data Fig. 8a). With five mice per group analysed, significant differences were shown between the histopathology score of the vehicle-treated group (1.2 out of 16) and the bruceine A-treated (0.2 out of 16) and gamabufotalin-treated (0.2 out of 16) groups (Extended Data Fig. 8b). These results suggest that the two natural compounds may provide therapeutic treatment for COVID-19.

Discussion

Several small-molecule inhibitors interfere with aspects of the life cycle of SARS-CoV-2, including the main protease Mpro^26,27,28,29 and the SARS-CoV-2 replicase RdRp³⁰. However, no potent small molecules at nanomolar range have been reported to inhibit the interaction of the viral S protein and host receptor ACE2. In the present study, we designed a brightly fluorescent PPI reporter that visualizes the interaction between spike protein and its receptor. This reporter is fluorogenic and reversible. Its large dynamic range and bright signal enable robust HTS of small-molecule inhibitors in living cells.

Using fluorogenic PPI reporters, we identified potent inhibitors that block the interaction of the two proteins in living cells at nanomolar concentrations, with verified inhibition of SARS-CoV-2 infection in both Vero cells and primary human airway cells at 15 nM. Additionally, the reversibility of our reporter allowed us to characterize the inhibition kinetics of the identified compounds which, interestingly, are proportional to their efficacy in cells. Our work demonstrates that this innovative reporter and structure-based engineering strategy can be applied to the future design of reporters for imaging key PPIs that mediate other virus–host interactions, and to screen for small-molecule inhibitors, following an approach similar to that developed here.

Our results suggest that the potent compounds bruceine A and gamabufotalin inhibit SARS-CoV-2 infection by blocking the S protein- and ACE2-mediated early stages of viral infection. First, inhibitors exert their activity on the early stages of viral infection because their antiviral activities are largely reduced when introduced at the later stages. Second, the SARS-CoV-2 binding assay showed that inhibitors block binding of the virus to host cells. Finally, using the pseudovirus, we further determined that the compounds exert their activity by blocking interaction of the S protein and ACE2 because the pseudovirus expresses only one of the SARS-CoV-2 proteins, the S protein. Furthermore, these compounds have no activity against SINV, which does not have the spike protein of SARS-CoV-2.

To gain insight into how these compounds inhibit S–ACE2 interaction, we docked both into potential pockets of the S–ACE2 protein complex and found that they were able to dock into the active site of ACE2 (Extended Data Fig. 9). An ACE2 inhibitor bound to the same site was reported to be effective in blocking SARS-CoV spike protein-mediated cell fusion³¹, demonstrating that the ACE2 inhibitor induces marked conformational changes in the active site that impacts surrounding residues, including those involved in binding to the spike protein. Considering the high sequence similarity between the two spike proteins (~75% homologous) of SARS-CoV and SARS-CoV-2, our compounds may also inhibit S–ACE2 interaction because the ACE2 inhibitor, possibly by conformational change of ACE2, blocks interaction with the spike protein.

Considering that the SARS-CoV-2 virus constantly changes through mutations, especially in the spike RBD domain, we tested the antiviral effect of bruceine A and gamabufotalin on Delta and Omicron variants. Both compounds maintained similar or slightly weaker inhibition effect on Delta and Omicron infection in Vero cells.

We demonstrate here that two natural compounds (gamabufotalin and bruceine A) exhibit potent antiviral efficacy in vivo when administered together with the virus as a prophylactic treatment. We chose these two inhibitors because their analogues showed largely reduced antiviral activity, indicating a structure–activity relationship. For example, cinobufotalin, which is structurally related to gamabufotalin, showed fivefold lower antiviral activity while bruceine D had fourfold lower antiviral activity than bruceine A (Fig. 3e). Both bruceine A and gamabufotalin reduced viral titre in the lung by between two and three logs at 3 days post-infection (d.p.i.) using K18-hACE2 mouse models of SARS-CoV-2 infection. These results were verified by immunofluorescence imaging of lung tissues. Furthermore, bruceine A eliminated coronavirus in mouse brain while gamabufotalin also significantly blocked SARS-CoV-2 replication in the brain. Histopathological analysis of lung showed a significant reduction in both alveolar and peribronchial inflammation. These findings demonstrate that both inhibitors significantly reduced SARS-CoV-2 replication in lung and brain, and prevented lung inflammation in the mouse model when the inhibitors were administered together with the virus. On the other hand, when the inhibitors were provided 24 h.p.i. they showed mild activity in blocking viral replication. This is consistent with results in cultured cells in which the inhibitors mainly blocked the early stages of viral infection. Our work suggests that the two compounds may be used as therapeutic treatment for COVID-19 via nasal application.

The PPI reporter we developed in this work visualizes the interaction between spike protein and its receptor ACE2. Our engineering strategy improved the reporter’s brightness by ~270-fold. The improved PPI reporter visualizes spike–ACE2 interaction with a wide dynamic range, which enabled HTS of small-molecule inhibitors of this PPI. This reporter strategy will have important applications in the future regarding the study of spike–receptor interactions, and it will also guide future engineering of reporters in imaging spike interaction with other receptors such as neuropillin-1 (NRP1). The spike protein of SARS-CoV-2 has been shown to interact with NRP1 (ref. 32), which facilitates infection of SARS-CoV-2 in neurons such as the NRP1-positive olfactory neurons facing the nasal cavity. We further show that the reporter is generalizable because it can be used to visualize the interaction between spike and NRP1 (Extended Data Fig. 10). This PPI reporter will be valuable in the study of spike–NRP1 interaction, as well as in identification of small-molecule inhibitors by HTS. Therefore, our work provides a general strategy in the engineering of PPI reporters for imaging interaction between a viral protein and its host cell receptor, and in screening for inhibitors that block viral interaction with host cells. Our PPI reporter, its engineering strategy and the screening approach will become a valuable resource in effectively dealing with any future pandemic by SARS-CoV-2 variants and other viruses.

Methods

Plasmid construction

To create the S^RBD::ΔACE2 SURF in pcDNA3.1, we used different truncations of the S protein and ACE2 and fused them to cSURF and nSURF, respectively. FKBP::Frb SURF was also cloned into pcDNA3.1, using full-length human FKBP CDS and human mTOR Frb domain (mTOR^2025–2114). pcDNA3.1-S^RBD-IFP2 and pcDNA3.1-IFP2-ΔACE2 were created by fusion of either S^RBD or ΔACE2 to IFP2 with a ten-amino-acid linker. All constructs were made using standard enzyme digestion and ligation methods.

Cell culture, transfection and live cell imaging

HEK293T cells (ATCC, no. 3216) were cultured in DMEM (Gibco, no. 11965-092) supplemented with 10% fetal bovine serum (FBS; Gibco, no. 16000044) and 1% penicillin/streptomycin/glutamine (Gibco Invitrogen). Vero-E6 cells (ATCC, no. 1586) were maintained in MEM (Gibco Invitrogen) supplemented with 10% FBS and 1% penicillin/streptomycin/glutamine. A549-ACE2 cells were a kind gift from P. Jackson’s laboratory at Stanford University. All cells were cultured at 37 °C in humidified 5% CO₂ incubators.

HEK293T cells were transfected using calcium phosphate or polyethylenimine. All live cell imaging was performed using a Nikon Eclipse Ti inverted microscope equipped with a Yokogawa CSU-W1 confocal scanner unit (Andor) and a CO₂/temperature control unit, using Nikon NIS-elements software.

For determination of inhibition kinetics of the identified drugs, we added each drug (10 μM) to HEK293T cells expressing S^RBD::ΔACE2 SURF, performed time-lapse imaging and analysed fluorescence changes over time. The inhibition kinetics of each drug were determined from normalized SURF fluorescence (by the red fluorescence of mCherry) over time. Seven drugs inhibited the PPI within 6 h, three of which (mitoxantrone (2 HCl), digoxin and cinobufotalin) inhibited it relatively rapidly with an inhibition half-life (τ) of ≤1.5 h.

HTS

A library of natural compounds containing 2,592 compounds (TargetMol, no. L6000) and an FDA-approved drug library (MedChemExpress, no. HY-L022) containing 1,516 compounds were used for HTS. For each screening, HEK293T cells were transfected with S^RBD::ΔACE2 SURF using polyethylenimine and seeded onto 96-well plates at ~5,000 cells per well, 6 h after transfection. One day after seeding, compounds were added to cells (5 or 10 μM final concentration for natural compounds or FDA-approved drugs, respectively). After 16 h of incubation with the compounds, images were taken for each well in the 488 and 561 channels. Total cell fluorescence was calculated using ImageJ. SURF fluorescence intensity was normalized by mCherry intensity and then compared with that of control wells incubated with DMSO to acquire fold-change values.

Cytotoxicity assay

Cytotoxicity of the identified compounds on Vero-E6 cells and HBEpCs was determined with WST-1 cell proliferation assays (Roche, no. 5015944001). For Vero-E6 cells, about 3,000 per well were seeded into 96-well plates and cultured for 1 day at 37 °C, followed by the addition of 1 μl of each compound at decreasing concentrations. After 18 or 72 h of incubation at 37 °C, WST-1 reagents were added to each well with incubation for 1 h. HBEpCs were cultured on 6.5 mm transwell inserts, compounds were added to the culture medium for 72 h then WST-1 reagent was added with incubation for 4 h. Absorbance at 450 and 650 nm was measured using a Tecan Infinite M1000 plate reader with iControl 1.8 software. All experiments were performed in three biological independent repeats.

SARS-CoV-2 strains including Delta and Omicron

A clinical isolate of SARS-CoV-2 (USA-WA1/2020, BEI NR-52281) was propagated in Vero-E6 cells and used for testing of drugs in this cell culture. A clinical isolate, SARS-CoV-2/human/USA/CA-UCSF-0001C/2020 (ref. 33), was propagated in A549 cells expressing ACE2 receptor under control of the CMV promoter and was used for both mouse and primary cell experiments.

Omicron and Delta variants were isolated in Vero-hACE2-TMPRSS2 cells (BEI NR-54970): 200 μl of nasal specimen was added to one well of a 96-well plate and serially diluted 1:1 over five additional wells; 100 μl of freshly trypsinized cells, resuspended in infection medium (DMEM with 10% FBS, 200 μg ml^–1 penicillin/streptomycin and 5 μg ml^–1 amphotericin B) at 2.5 × 10⁵ cells ml^–1, was then added to each sample dilution. Cells were cultured at 37 °C with 5% CO₂ and checked for cytopathic effects at 2–3 d.p.i. Supernatants were harvested on day 3 after inoculation; 200 μl of P0 was used to infect a confluent T25 flask to generate a P1 culture then harvested after 3 days. Viruses were passaged three times after isolation to obtain a sufficiently high titre (>10⁷ PFU ml^–1). Viral titres were quantified by plaque assay. All infection experiments were performed at biosafety level 3.

Antiviral assay on Vero-E6 cells

For assessment of initial antiviral activity and EC₅₀ of drugs, 70% confluent monolayers of Vero-E6 cells (2 × 10⁵ per well in 24-well plates) were pretreated with drug for 3 h (pretreatment) and then infected with SARS-CoV-2 (MOI = 0.5) at 37 °C for 1 h. Virus solution was removed, then cells were washed twice with PBS and added to fresh medium containing drug/natural compound or DMSO to continue culture. At 16 h.p.i., viral titres of supernatants were determined by plaque assay.

Plaque assay

Confluent monolayers of Vero-E6 cells grown in six-well plates were incubated with serial dilutions of virus samples (250 μl per well) at 37 °C for 1 h. Next, cells were overlaid with 1% agarose (Invitrogen) prepared with MEM supplemented with 2% FBS and 1% penicillin/streptomycin/glutamine (100x, Gibco Invitrogen). After 3 days cells were fixed with 4% formaldehyde for 2 h, the overlay was discarded and samples stained with crystal violet dye.

Adsorption assay

Vero-E6 cell monolayers at 70% confluence grown on coverslips were preincubated with natural compounds at 37 °C for 3 h. Cells were then added to SARS-CoV-2 virus at MOI = 100 and immediately transferred to 4 °C for 1 h. The virus suspension was quickly removed and cells were washed three times with ice-cold PBS. Cells were immediately fixed with 4% PFA for 30 min at room temperature. PFA was then washed away with PBS and quenched with 1 M glycine in PBS.

Drug staging experiment

Drugs were added at different stages of SARS-CoV-2 infection. For −4 h samples, Vero-E6 cells (~2 × 10⁵ per well in 24-well plates) were pretreated with 2 μM compound for 3 h (preincubation) and infected with SARS-CoV-2 at MOI = 5. After 1 h of incubation at 37 °C, cells were washed twice with PBS to remove virus and then incubated with fresh medium containing drug; for +1 h and +4 h samples, there was no preincubation, and drugs were added 1 or 4 h.p.i., respectively. At 16 h.p.i., supernatant was collected for plaque assay and cells harvested using Trizol reagent (Ambion) for RT–qPCR.

Pseudovirus rVSV-eGFP-SARS-CoV-2

We preincubated ACE2-expressing A549 cells with compounds for 3 h, infected the cells with rVSV-eGFP-SARS-CoV-2 (MOI = 1) at 37 °C for 1 h then removed the pseudovirus, added compounds again and cultured the cells for 16 h. Cells were then imaged and infection rates calculated based on percentage of cells with eGFP signal.

SINV–eGFP generation

The eGFP expressing Sindbis virus (SINV–eGFP) was cloned by insertion of the eGFP coding sequence into the XbaI site of the double-subgenomic Sindbis vector pTE392J (provided by C. Rice). Viral genomic RNA was obtained by in vitro transcription using MEGAscript SP6 kit (Invitrogen) with transfection into BHK-21 cells. The initial virus stock was collected from the supernatant of transfected cells and amplified once by infecting a new culture of BHK-21 cells (passage 1, P1). Finally, P1 viral stock titration was performed by plaque assay in BHK-21 cells. A procedure similar to that for SARS-CoV-2 was followed here to examine effects of the natural compounds on SINV.

Molecular docking

The structure of SARS-CoV-2 S protein complexed with human ACE2 (PDB ID: 6VW1) was used in molecular docking calculations, with calculations performed as previously described³⁴. Protonation states of residues in the ACE2 active site were visually checked. Atom charges were assigned to the structure using the AMBER united atom force field. Energy grids were pregenerated with CHEMGRID for van der Waals potential, with QNIFFT for Poisson–Boltzmann-based electrostatic potential and with SOLVMAP for ligand desolvation penalty. Presampling three-dimensional conformer libraries were generated by ZINC. Compounds were docked against the ACE2 active site using DOCK3.7. Docking poses were outputted and examined visually. Figures were generated using UCSF Chimera.

Infection of primary cells

Primary human nasal epithelial cells (PromoCell) were pretreated with compound for 3 h, infected with virus at MOI = 3, incubated with medium containing 2 μM inhibitor for 48 h, then fixed with 4% paraformaldehyde in PBS.

Primary HBEpCs were cultured on 6.5 mm transwell inserts (Corning) for 4 weeks at an air–liquid interface to obtain well-differentiated epithelium. Mucus was removed immediately before viral infection by washing the apical surface with a prewarmed solution of 10 mM dithiothreitol (Thermo) in PBS for 10 min, then washed twice with PBS without dithioreitol. Cells were inoculated by the addition of virus to the apical surface at MOI = 1. After 1 h, the apical surface was washed twice with PBS and cells were returned to the incubator. Cells were fixed at 72 h.p.i. in 4% paraformaldehyde in PBS.

Immunofluorescence

Both Vero-E6 and primary cells were fixed in 4% paraformaldehyde, permeabilized by PBST (PBS with 0.1% Triton X-100), blocked with 2% BSA and 10% goat serum and incubated overnight at 4 °C with antinucleocapsid antibody (1:500; Genetex, no. GTX135357) and/or antispike antibody (1:600; Genetex, no. GTX632604). After washing three times with PBST, cells were incubated with goat antirabbit IgG H&L (1:800, Alexa Fluor 488; Abcam, no. ab150077) or goat antimouse IgG H&L (1:800, Alexa Fluor 555; Abcam, no. ab150114).

For both Vero-E6 and primary HBEpCs, images were acquired using a Nikon Eclipse Ti-E Spinning Disk microscope with a ×60 numerical aperture (NA) 1.49 Plan-Apo objective. The software used was NIS-Element AR 4.20. For primary HNEpCs, images were acquired on an Everest deconvolution workstation (Intelligent Imaging Innovations) equipped with a Zeiss AxioImager Z1 microscope, CoolSnapHQ cooled CCD camera (Roper Scientific) and a ×40 NA 1.3 Plan-Apochromat objective lens (Zeiss, no. 420762-9800). The software used for acquisition was SlideBook (v.6). All images were processed in ImageJ. Fluorescence signals were calculated by pixel intensity using Analyze Particle function in ImageJ, and cell numbers calculated using Cell Counter plugin in ImageJ.

Three wells of either Vero-E6 or primary cell cultures were infected for each condition. Ten field views for each well of Vero-E6, three for each HNE culture and ten for each HBE culture were imaged. In total, 1,500 cells were analysed for each treatment for Vero-E6 or HBEpCs, and 500 for each treatment for HNEpCs. Total N- and S-positive cell numbers per 100 cells were used for statistical analysis.

vRNA isolation and qPCR

Genomic RNA was extracted from cell pellets using Trizol (Ambion) according to the manufacturer’s instructions. RNA was treated with DNAse I (NEB) and used as a template to reverse transcribe complementary DNA with the Iscript kit (NEB). qPCR was performed using the Luna Universal qPCR Master Mix (NEB) and a CFX connect qPCR Detection System (Bio-Rad). To determine the number of vRNA copies per millilitre, plasmids containing the nucleocapsid gene of SARS-CoV-2 (cloned from the USA-WA1/2020 isolate) were used as standards and diluted serially tenfold to determine target copy numbers. Threshold cycle values were plotted against the number of target copies, with the resultant standard curve used to determine the number of genome equivalents of vRNA in samples. For cell pellet samples, vRNA copy number was normalized to the housekeeping gene HRT1. All samples were within the range of linearity of a standard curve, and primer efficiency was 100 ± 5%. The primer sequences targeting nucleocapsid were 5′-TCCTGGTGATTCTTCTTCAGG-3′ and 5′-TCTGAGAGAGGGTCAAGTGC-3′; HRT1 primer sequences were 5′-GGTCCTTTTCACCAGCAAGCT-3′ and 5′-TGACACTGGCAAAACAATGCA-3′.

Inhibitor preparation for in vivo study

Natural products were obtained from MedChemExpress (bruceine A, no. HY-N0841 and gamabufotalin, no. HY-N0883). Natural products were prepared at 100 mg ml^–1 in DMSO, then each was diluted in solvent ((5% Tween 80; Sigma, no. P4780) and 40% PEG300 (MedChemExpress, no. HY-Y0873, 55% PBS)).

Toxicity testing in animals

Natural products were delivered intranasally. Briefly, C57BL6/J mice were anaesthetized with isoflurane and inoculated with solvent and compounds at different diluted concentrations in 50 μl (6.50, 3.25 and 1.63 mg kg^–1). Mice were treated every 24 h for 3 days. No obvious toxicity was observed at dosages up to 6.5 mg kg^–1 per mouse by intranasal inoculation.

Antiviral effect testing in mouse models

All antiviral studies were performed in an animal biosafety level 3 facility at the Department of Immunology and Microbiology, University of California, San Francisco (UCSF). SARS-CoV-2 experiments were performed in 8–10-week-old K18-hACE2 mice (The Jackson Laboratory, no. 034860, B6.Cg-Tg (K18-hACE2) 2Prlmn/J, hemizygous) and C57BL6/J mice (The Jackson Laboratory, no. 000664, RRID:IMSR_JAX:000664). Mice were housed and bred in a specific-pathogen-free, AAALAC-certified animal facility at UCSF at ~23 °C and 50% humidity under a 12/12-h light/dark cycle, with standard chow diet provided. All procedures were performed in accordance with the guidelines of the Laboratory Animal Center of the National Institutes of Health. The Institutional Animal Care and Use Committee of the UCSF approved all animal protocols (approved protocol no. AN178420-02B).

For K18-hACE2 mice, on day 0 we administered solvent, bruceine A and gamabufotalin and, simultaneously, 1,000 PFU of SARS-CoV-2 in 50 μl (25 µl of virus solution in PBS plus 25 µl of drug solution or solvent) intranasally. Bruceine A and gamabufotalin were then administrated intranasally once per day for 2 days. Mice were anaesthetized with isoflurane before intranasal infection. This experiment was repeated twice independently; a total of eight mice were used in each control or drug-treated group.

For C57BL6/J mice, on day 0 we administered 10⁵ PFU of SARS-CoV-2 (B.1.1.7) intranasally. Bruceine A and gamabufotalin were then administrated intranasally at 24 and 48 h.p.i. A total of five mice were used in each control or drug-treated group.

For virus tissue distribution, at 3 d.p.i. animals were humanely euthanized and lungs and brains harvested. Half of the lungs were fixed in 4% PFA for 72 h, transferred to 70% ethyl alcohol, embedded in paraffin and processed for H&E and immunofluorescence staining. The other half of the lungs, and brains, were weighted and homogenized in 1 ml of 2% FBS MEM medium with gentleMACS C Tubes (Miltenyi Biotec, no. 130-093-237). Samples were centrifuged at 2,500g for 10 min and the supernatant transferred to a new set of tubes, then frozen at −80 °C for viral titration by plaque assay. Infectious titres were quantified by plaque assay titration using Vero-E6 cells as described above. Unpaired t-testing was used for tissue distribution analysis.

Mouse lung histological analysis

Tissue samples were embedded in paraffin. Lung tissue blocks were cut into 5 μm sections in the Gladstone pathology and light microscope core facility. Sections were stained with H&E, imaged under digital light microscopy in the UCSF Nikon imaging center and then analysed by ImageJ. H&E-stained lung sections were examined using a semiquantitative method previously described³⁵. The severity of four histopathological parameters—(1) perivascular inflammation, (2) bronchial or bronchiolar epithelial degeneration or necrosis, (3) bronchial or bronchiolar inflammation and (4) alveolar inflammation)—was defined and a score assigned for each parameter (0, none; 1, <25%; 2, 26–50%; 3, 51–75%; and 4, >75%). Unpaired t-testing was used for statistical analysis.

Immunofluorescence examination of lung sections

Deparaffinization, rehydration and heat-induced epitope retrieval were performed on an ST4020 small linear stainer (Leica Biosystems). For deparaffinization, slides were baked at 70 °C for 1.0–1.5 h followed by rehydration in descending concentrations of ethanol (100% twice, 95% twice and 80%, 70% and ddH₂O twice; each step for 30 s). Washes were performed using a Leica ST4020 Linear Stainer (Leica Biosystems) programmed to three dips per wash for 30 s each. Heat-induced epitope retrieval was performed in a Lab VisionTM PT module (Thermo Fisher) using Dako Target Retrieval Solution pH 9.0 (no. S236784-2, DAKO Agilent) at 97 °C for 10 min and cooled to 65 °C. After further cooling to room temperature for 30 min, slides were washed three times (10 min each) in Tris-buffered saline (TBS) containing 0.1% Tween 20 (TBST, Cell Marque). Sections were then blocked in 5% normal donkey serum in TBST at room temperature for 1 h, followed by incubation with primary antibodies in the blocking solution. After overnight incubation of primary antibodies at 4 °C, sections were washed three times with TBST and stained with the appropriate secondary antibodies in PBS plus 3% bovine serum albumin, 0.4% saponin and 0.02% sodium azide at room temperature for 1 h. Next, sections were washed three times with TBST and mounted with ProLong Gold Antifade mounting medium containing DAPI (Invitrogen). The primary antibodies and final titrations used were mouse antiacetylated α-tubulin (1:300; Santa Cruz, no. sc-23950), rabbit anti-SARS-CoV-2 nucleocapsid (1:1,000; GeneTex, no. GTX135361) and mouse anti-SARS-CoV-2 spike (1:600; GeneTex, no. GTX632604). Secondary antibodies included highly cross-adsorbed donkey antirabbit Alexa Fluor Plus 647 (1:500; Thermo, no. A32795) and highly cross-adsorbed donkey antimouse Alexa Fluor Plus 555 (1:500; Thermo, no. A32773). Fluorescence-immunolabelled images were acquired using either a Zeiss AxioImager Z1 or Keyence BZ-X710 fluorescent microscope. Post-imaging processing was performed using the FIJI package ImageJ 1.52t.

Statistics and reproducibility

All statistics were performed in GraphPad Prism or Microsoft Excel. The IC₅₀ or EC₅₀ of drug inhibition, cytotoxicity and antiviral activity was calculated using the nonlinear fit function (variable slope). All SURF reporter testing and cell immunofluorescence staining were repeated at least three times. Mouse lung section immunofluorescence staining was performed in two mice per group, with three slides per mouse analysed. H&E staining was performed in two mice per group for K18-hACE2 mice, and five mice per group for C57BL6 mice; three slides per mouse were prepared. No statistical methods were used to predetermine sample size, but our sample sizes are similar to those reported in previous publications^29,35. Cell cultures and animals were all randomly grouped. Library screening and initial antiviral tests (Figs. 2 and 3) were carried out blindly, labelled with numbers but not compound names. The following experiments were not blind, since differences between compound and control groups were readily observed from imaging data or antiviral tests. No more than two data points, if any, for one drug in drug library screening were excluded due to low imaging quality. One outlier data point in the RT–qPCR assay for the anti-Omicron variant test, DMSO group was excluded due to low RNA yield.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data are available within the paper and its Supplementary information. The complete screening results and raw data for mouse experiments can be found in the Source Data section.

References

Taylor, P. C. et al. Neutralizing monoclonal antibodies for treatment of COVID-19. Nat. Rev. Immunol. 21, 382–393 (2021).
Article CAS Google Scholar
Dai, L. & Gao, G. F. Viral targets for vaccines against COVID-19. Nat. Rev. Immunol. 21, 73–82 (2021).
Article CAS Google Scholar
Annavajhala, M. K. et al. Emergence and expansion of SARS-CoV-2 B.1.526 after identification in New York. Nature 597, 703–708 (2021).
Article CAS Google Scholar
Lan, J. et al. Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor. Nature 581, 215–220 (2020).
Article CAS Google Scholar
Wang, Q. et al. Structural and functional basis of SARS-CoV-2 entry by using human ACE2. Cell 181, 894–904 (2020).
Article CAS Google Scholar
Wrapp, D. et al. Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation. Science 367, 1260–1263 (2020).
Article CAS Google Scholar
Yan, R. et al. Structural basis for the recognition of the SARS-CoV-2 by full-length human ACE2. Science 367, 1444–1448 (2020).
Article CAS Google Scholar
Glebov, O. O. Understanding SARS‐CoV‐2 endocytosis for COVID‐19 drug repurposing. FEBS J. 92, 207–208 (2020).
Google Scholar
Shang, J. et al. Cell entry mechanisms of SARS-CoV-2. Proc. Natl Acad. Sci. USA 117, 11727–11734 (2020).
Article CAS Google Scholar
Cao, L. et al. De novo design of picomolar SARS-CoV-2 miniprotein inhibitors. Science 370, 426–431 (2020).
Article CAS Google Scholar
Chan, K. K. et al. Engineering human ACE2 to optimize binding to the spike protein of SARS coronavirus 2. Science 369, 1261–1265 (2020).
Article CAS Google Scholar
Yuan, M. et al. A highly conserved cryptic epitope in the receptor-binding domains of SARS-CoV-2 and SARS-CoV. Science 368, 630–633 (2020).
Article CAS Google Scholar
Zhang, G., Pomplun, S., Loftis, A. R., Loas, A. & Pentelute, B. L. The first-in-class peptide binder to the SARS-CoV-2 spike protein. Preprint at bioRxiv https://doi.org/10.1101/2020.03.19.999318 (2020).
Kumagai, A. et al. A bilirubin-inducible fluorescent protein from eel muscle. Cell 153, 1602–1611 (2013).
Article CAS Google Scholar
To, T.-L., Zhang, Q. & Shu, X. Structure-guided design of a reversible fluorogenic reporter of protein-protein interactions. Protein Sci. 25, 748–753 (2016).
Article CAS Google Scholar
Szymczak, A. L. et al. Correction of multi-gene deficiency in vivo using a single ‘self-cleaving’ 2A peptide-based retroviral vector. Nat. Biotechnol. 22, 589–594 (2004).
Article CAS Google Scholar
Campbell, R. E. et al. A monomeric red fluorescent protein. Proc. Natl Acad. Sci. USA 99, 7877–7882 (2002).
Article CAS Google Scholar
Shaner, N. C. et al. Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat. Biotechnol. 22, 1567–1572 (2004).
Article CAS Google Scholar
Shu, X., Shaner, N. C., Yarbrough, C. A., Tsien, R. Y. & Remington, S. J. Novel chromophores and buried charges control color in mFruits. Biochemistry 45, 9639–9647 (2006).
Article CAS Google Scholar
Shu, X. et al. Mammalian expression of infrared fluorescent proteins engineered from a bacterial phytochrome. Science 324, 804–807 (2009).
Article Google Scholar
Yu, D. et al. An improved monomeric infrared fluorescent protein for neuronal and tumour brain imaging. Nat. Commun. 5, 3626 (2014).
Article CAS Google Scholar
Zhang, J., Chung, T. & Oldenburg, K. A simple statistical parameter for use in evaluation and validation of high throughput screening assays. J. Biomol. Screen. 4, 67–73 (1999).
Article CAS Google Scholar
Case, J. B. et al. Neutralizing antibody and soluble ACE2 inhibition of a replication-competent VSV-SARS-CoV-2 and a clinical isolate of SARS-CoV-2. Cell Host Microbe 28, 475–485 (2020).
Article CAS Google Scholar
Ryu, G. & Shin, H.-W. SARS-CoV-2 infection of airway epithelial cells. Immune Netw. 21, e3 (2021).
Article Google Scholar
Winkler, E. S. et al. SARS-CoV-2 infection of human ACE2-transgenic mice causes severe lung inflammation and impaired function. Nat. Immunol. 21, 1327–1335 (2020).
Article CAS Google Scholar
Dai, W. et al. Structure-based design of antiviral drug candidates targeting the SARS-CoV-2 main protease. Science 368, 1331–1335 (2020).
Article CAS Google Scholar
Jin, Z. et al. Structure of Mpro from COVID-19 virus and discovery of its inhibitors. Nature 582, 289–293 (2020).
Article CAS Google Scholar
Zhang, L. et al. Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors. Science 368, 409–412 (2020).
Article CAS Google Scholar
Li, X. et al. Ethacridine inhibits SARS-CoV-2 by inactivating viral particles. PLoS Pathog. 17, e1009898 (2021).
Article CAS Google Scholar
Yin, W. et al. Structural basis for inhibition of the RNA-dependent RNA polymerase from SARS-CoV-2 by remdesivir. Science 368, 1499–1504 (2020).
Article CAS Google Scholar
Huentelman, M. J. et al. Structure-based discovery of a novel angiotensin-converting enzyme 2 inhibitor. Hypertension 44, 903–906 (2004).
Article CAS Google Scholar
Daly, J. L. et al. Neuropilin-1 is a host factor for SARS-CoV-2 infection. Science 370, 861–865 (2020).
Article CAS Google Scholar
Samuel, R. M. et al. Androgen signaling regulates SARS-CoV-2 receptor levels and is associated with severe COVID-19 symptoms in men. Cell Stem Cell 27, 876–889 (2020).
Article CAS Google Scholar
Bender, B. J. et al. A practical guide to large-scale docking. Nat. Protoc. 16, 4799–4832 (2021).
Article CAS Google Scholar
White, K. M. et al. Plitidepsin has potent preclinical efficacy against SARS-CoV-2 by targeting the host protein eEF1A. Science 371, 926–931 (2021).
Article CAS Google Scholar

Download references

Acknowledgements

We thank D. Gordon and N. Krogan for sharing the cDNA of SARS-CoV-2 spike protein; S. P. J. Whelan for sharing the rVSV-eGFP-SARS-CoV-2 pseudovirus; S. G. Remesh, K. Leung and J. Wells for help with the the PPI assay; and L. Rodriguez for help with the plaque assay for HBEpCs. This work was supported by the UCSF Program for Breakthrough Biomedical Research (funded in part by the Sandler Foundation), NIH (no. R35GM131766 to X.S. and nos. R01 AI36178, AI40085 and P01 AI091575), the Bill and Melinda Gates Foundation and the DARPA Intercept programme (contract no. HR0011-17-2-0027 to R.A.) and NIH (nos. R35 HL145235 and U19 AI077439 to D.J.E.)

Author information

These authors contributed equally: Junjiao Yang, Yinghong Xiao, Peter V. Lidsky.

Authors and Affiliations

Department of Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, CA, USA
Junjiao Yang, Shiming Peng, Chan-I Chung, Xiaoquan Li, Brian K. Shoichet & Xiaokun Shu
Cardiovascular Research Institute, University of California, San Francisco, San Francisco, CA, USA
Junjiao Yang, Chan-I Chung, Xiaoquan Li & Xiaokun Shu
Department of Microbiology and Immunology, University of California, San Francisco, San Francisco, CA, USA
Yinghong Xiao, Peter V. Lidsky, Miguel A. Garcia-Knight, Michel Tassetto & Raul Andino
Department of Baxter Laboratory for Stem Cell Biology, Department of Microbiology & Immunology, Stanford University, Stanford, CA, USA
Chien-Ting Wu & Peter K. Jackson
Lung Biology Center, Department of Medicine, University of California, San Francisco, San Francisco, CA, USA
Luke R. Bonser, Khadija Ghias, Kirsten L. Hargett & David J. Erle
Department of Otolaryngology–Head and Neck Surgery, Stanford University, Stanford, CA, USA
Tsuguhisa Nakayama & Jayakar V. Nayak
Division of Allergy, Immunology, and Rheumatology, Department of Pediatrics, Stanford University School of Medicine, Stanford, CA, USA
Ivan T. Lee
Department of Otolaryngology–Head and Neck Surgery, VA Palo Alto Health Care System, Palo Alto, CA, USA
Jayakar V. Nayak

Authors

Junjiao Yang
View author publications
You can also search for this author in PubMed Google Scholar
Yinghong Xiao
View author publications
You can also search for this author in PubMed Google Scholar
Peter V. Lidsky
View author publications
You can also search for this author in PubMed Google Scholar
Chien-Ting Wu
View author publications
You can also search for this author in PubMed Google Scholar
Luke R. Bonser
View author publications
You can also search for this author in PubMed Google Scholar
Shiming Peng
View author publications
You can also search for this author in PubMed Google Scholar
Miguel A. Garcia-Knight
View author publications
You can also search for this author in PubMed Google Scholar
Michel Tassetto
View author publications
You can also search for this author in PubMed Google Scholar
Chan-I Chung
View author publications
You can also search for this author in PubMed Google Scholar
Xiaoquan Li
View author publications
You can also search for this author in PubMed Google Scholar
Tsuguhisa Nakayama
View author publications
You can also search for this author in PubMed Google Scholar
Ivan T. Lee
View author publications
You can also search for this author in PubMed Google Scholar
Jayakar V. Nayak
View author publications
You can also search for this author in PubMed Google Scholar
Khadija Ghias
View author publications
You can also search for this author in PubMed Google Scholar
Kirsten L. Hargett
View author publications
You can also search for this author in PubMed Google Scholar
Brian K. Shoichet
View author publications
You can also search for this author in PubMed Google Scholar
David J. Erle
View author publications
You can also search for this author in PubMed Google Scholar
Peter K. Jackson
View author publications
You can also search for this author in PubMed Google Scholar
Raul Andino
View author publications
You can also search for this author in PubMed Google Scholar
Xiaokun Shu
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

J.Y. and X.S. initiated the project and designed fluorescent reporters and assays. J.Y., C-I.C. and X.L. prepared the drug and natural compound library. J.Y. performed screens and characterized the kinetics of PPI inhibition in cells. Y.X., P.V.L. and R.A. designed SARS-CoV-2 biosafety level 3 experiments in cells and animals and interpreted data. Y.X. and P.V.L. performed plaque assays and calculated plaque numbers. J.Y., Y.X. and P.V.L. performed quantitative analysis and plotted data. M.A.G.-K. isolated virus variants, performed RT–qPCR and analysed data. J.Y. performed toxicity assays. P.V.L. and Y.X. infected Vero cells. J.Y. performed immunostaining and analysed immunofluorescence. Y.X., P.V.L., T.N., J.V.N., C.-T.W., P.K.J. and R.A. designed HNEpC infection experiments. Y.X. and P.L. performed viral infection. C.-T.W., T.N. and I.T.L. performed cell culturing. C.-T.W. conducted immunostaining and imaging. J.Y. and X.L. performed quantitative analysis. J.Y. and L.R.B. designed HBEpC infection experiments, L.R.B. and K.G. performed HBEpC culture. P.V.L. and M.T. performed viral infection. J.Y. performed immunostaining, imaging and quantitative analysis. S.P. and B.K.S. performed and analysed molecular docking. X.S., J.Y. and Y.X. wrote the manuscript. J.Y., Y.X., P.V.L., M.K., C.-T.W., S.P. and M.T. wrote the methods used. All authors edited and contributed to the final draft.

Corresponding authors

Correspondence to Raul Andino or Xiaokun Shu.

Ethics declarations

Competing interests

The authors declare the following competing interest: J.Y. and X.S. have filed a patent on the imaging assay.

Peer review

Peer review information

Extended Data Fig. 10 The PPI reporter SURF can be used to image interaction between spike and NRP1.

Spike (529-685aa) and NRP1 (273-429aa) are fused to each part of SURF. The SURF reporter was transiently co-transfected into HEK293T with mApple_T2A_H2B-mIFP. Green: SURF fluorescence. Red: mCherry fluorescence. Blue: mIFP fluorescence. Scale bar: 20 μm. The experiments were repeated for three times independently with similar results.

Source data

Supplementary information

Supplementary Information

Supplementary Figs. 1–12 with legends.

Reporting Summary

Peer Review File

Supplementary Data

Statistical Source Data for Supplementary Figures.

Reprints and permissions

About this article

Cite this article

Yang, J., Xiao, Y., Lidsky, P.V. et al. Fluorogenic reporter enables identification of compounds that inhibit SARS-CoV-2. Nat Microbiol 8, 121–134 (2023). https://doi.org/10.1038/s41564-022-01288-5

Download citation

Received: 03 February 2022
Accepted: 16 November 2022
Published: 05 January 2023
Issue Date: January 2023
DOI: https://doi.org/10.1038/s41564-022-01288-5

This article is cited by

SURFing SARS-CoV-2 inhibition
- Agustina Taglialegna
Nature Reviews Microbiology (2023)