Identification of natural compounds as SARS-CoV-2 entry inhibitors by molecular docking-based virtual screening with bio-layer interferometry

Abstract

Coronavirus Disease 2019 (COVID-19) is caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), which enter the host cells through the interaction between its receptor binding domain (RBD) of spike glycoprotein with angiotensin-converting enzyme 2 (ACE2) receptor on the plasma membrane of host cell. Neutralizing antibodies and peptide binders of RBD can block viral infection, however, the concern of accessibility and affordability of viral infection inhibitors has been raised. Here, we report the identification of natural compounds as potential SARS-CoV-2 entry inhibitors using the molecular docking-based virtual screening coupled with bilayer interferometry (BLI). From a library of 1871 natural compounds, epigallocatechin gallate (EGCG), 20(R)-ginsenoside Rg3 (RRg3), 20(S)-ginsenoside Rg3 (SRg3), isobavachalcone (Ibvc), isochlorogenic A (IscA) and bakuchiol (Bkc) effectively inhibited pseudovirus entry at concentrations up to 100 μM. Among these compounds, four compounds, EGCG, Ibvc, salvianolic acid A (SalA), and isoliensinine (Isl), were effective in inhibiting SARS-CoV-2-induced cytopathic effect and plaque formation in Vero E6 cells. The EGCG was further validated with no observable animal toxicity and certain antiviral effect against SARS-CoV-2 pseudovirus mutants (D614G, N501Y, N439K & Y453F). Interestingly, EGCG, Bkc and Ibvc bind to ACE2 receptor in BLI assay, suggesting a dual binding to RBD and ACE2. Current findings shed some insight into identifications and validations of SARS-CoV-2 entry inhibitors from natural compounds.

Keywords: Bio-layer interferometry; Competitive binding; Molecular docking; Molecular dynamics simulation; Natural compounds; SARS-CoV-2.

Graphical abstract

Fig. 1

Identified natural compounds directly interact with RBD protein by BLI binding kinetics assay. (A) Immobilization of His-tagged ACE2 to Ni-NTA probe. (B) Binding affinity of RBD to ACE2. KD is calculated by fitting of maximum BLI response at different concentrations to a sigmoidal curve. (C) Loading of RBD to SSA probe in association conditions. (D) Schematic illustration of the BLI binding kinetics assay. (E) BLI sensorgrams structures of 14 compounds showing their binding to RBD. Each compound was examined in BLI screening from compound library and a new batch for validation of reproducibility. For each compound, sensorgram of the test with better result is shown. Chemical structures of compounds are displayed.

Fig. 2

The natural small-molecules inhibit RBD protein binding to hACE2 by ELISA and ICC assays. (A) ELISA results of 14 compounds. EGCG, IscA, SalA, Isl, Bvc and Bvcn had high activities. In addition, Ibvc, Bkc, CvbD and CalB were weakly active. Ceph, TimA, SRg3 and RRg3 were almost inactive. Data are presented as mean ± SD. Data are analysed with one-way ANOVA (*P < 0.05; **P < 0.01; ***P < 0.005). (B) Visualization of Spike-RBD binding on human ACE2 receptor in HEK293 cells. Representative image of RBD red fluorescence on HEK293 cells in the presence of compounds.

Fig. 3

The anti-viral effect of natural small-molecules in hACE2 overexpressing HEK293 cells infected by SARS-CoV-2-pseudotyped lentivirus. Representative fluorescence microscopic images of cells expressing ACE2-mCherry, either treated with EGFP-pseudovirus or treated with EGFP-pseudovirus in the presence of 100 μM compounds, are shown; scale bars represent 48.5 µm (upper). Data of pseudovirus entry inhibition at different concentrations are standardized to the mean of fluorescence intensity at 0 μM concentration and presented as mean ± S.D. of three replicates; EC₅₀ is calculated by fitting a sigmoidal curve in GraphPad Prism 8.0 software (lower).

Fig. 4

The anti-viral effect of natural small-molecules toward live SARS-CoV-2 virus. (A) Inhibitory effects of compounds on CPE effect, such as confluence, number of cells and morphology, induced by authentic SARS-CoV-2 infection in Vero E6 cell culture and cytotoxicity of compounds. (B, C) Compounds effectively inhibit plaque-forming ability of SARS-CoV-2.

Fig. 5

EGCG inhibits the infection of 4 mutant S-pseudotyped lentivirus in human ACE2 overexpressing cells. HEK293 cells were transiently transfected with hACE2-mCherry (red). After 24 h, the hACE2 overexpressing cells were infected by RBD wild-type and mutant S-pseudotyped lentivirus (D614G, N501Y, N439K & Y453F) (green) in the presence of 0–100 μM EGCG. The infected cells were then replaced with fresh medium and continually incubated for 48 h. All images were captured by confocal microscopy using a Leica SP8 (×40 oil immersion objective lens). Data of pseudovirus entry inhibition at different concentrations are standardized to the mean of fluorescence intensity (quantified by Image J) at 0 μM concentration and presented as mean ± S.D. of three replicates; *P < 0.05, **P < 0.01, student-t-test analysis.

Fig. 6

Computational modeling experiment to predict binding sites of natural compounds. (A) Possible binding sites identified by F-pocket. (B-D) Active compounds clusters into three binding sites (B) P1; (C) P2 and (D) P3. (E) Predicted binding pockets. P1 & P2 are located at the core domain of RBD; P3 is located at the distal loop of the RBM. (F-H) Equilibration in 10 ns of all compounds in their respective pockets confirms binding stability.

Fig. 7

Validation of RBD binding pockets with BLI competitive binding assay. (A) Non-competitive interaction. (B) Weakly competitive interaction. (C) Competitive interaction between primary and second compounds in BLI cross-competitive binding assay. (D) Example of a compound blocking its own binding at high concentration. (E) Heatmap matrix representing the results of BLI-based competitive binding experiment of 8 compounds as indicated. Value in each cell represents the percent reduction of the signal of the second compound in the presence of high concentration of the primary compound. It is calculated as signal of mixed compound/signal of second compound alone ×100%, maximizing at 100% and minimizing at 0. Non-competition rate of an interaction was calculated as 100%×|z/y| with a maximum of 100% and minimum of 0. Non-competition rate of ≥ 60% indicates that binding of second compound is not significantly diminished by previous first compound binding; Non-competition rate of ≤ 20% is considered to constitute competition. 0.2