Preparation of the target protein and ligand

The ligand calceolarioside B was downloaded from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) [19]. The protein 7UB5, representing the SARS-CoV-2 Omicron BA.2 S protein trimer with the three-receptor-binding domain (RBD) configuration and lacking the P986-P987 stabilizing mutations (designated as S-GSAS-Omicron BA.2), was retrieved from the RCSB Protein Data Bank (https://www.rcsb.org/) [20]. The Schrödinger LigPrep tool (https://www.schrodinger.com/) was used to prepare a high-quality ligand for further molecular docking. Protein preparation, including preprocessing, optimization, water removal, and minimization, were undertaken using the Schrödinger Protein Preparation Wizard (https://www.schrodinger.com/) [21] SiteMap tools were used to evaluate the potential protein binding sites. A receptor grid was generated by applying the SiteMap results to Receptor grid generation in the Grid based Ligand Docking with Energetics (Glide) application, (version 9.1, Schrödinger, https://www.schrodinger.com/) of Maestro version 12.8 (Schrödinger). After screening, the receptor grid located in the RBD was used for molecular docking.

Molecular docking

The ligand is docked to the protein (7UB5) by employing the docking protocol of Glide version 9.1 with the receptor grid screened from the preparation process. After docking, each of the binding poses was then used in the Calculate Energy module of the MacroModel application (BatchMin V13.2; Schrödinger) to calculate the total energy.

Molecular dynamics simulation

The molecular dynamics of the interaction between calceolarioside B and SARS-CoV-2 Omicron BA.2 RBD were estimated using the GROMACS software (version 2022.2, https://manual.gromacs.org/2022.2/download.html) [22]. The topologies of RBD and calceolarioside B were generated using the OPSL-AA force field and LigParGen server (https://traken.chem.yale.edu/ligpargen/) [23], respectively. RBD and calceolarioside B complexes were constructed and then subjected to a dodecahedral water box SPC water model (https://www.gromacs.org/) [24]. The system was neutralized after adding suitable counter ions. Energy minimization was performed with a force constant of 1000 kJ mol−1 nm−1 and a maximum of 50,000 steps. During equilibration, NVT equilibration was set at 300 K and NPT at 1 bar. The molecular dynamics simulation was set to 1 ns. The root-mean-squared deviation (RMSD) and interaction energy were also ascertained by the GROMACS software.

Biolayer interferometry (BLI) assay

SARS-CoV-2 Omicron BA.2 Spike-RBD protein (His Tag) (Cat: 40592-V08H123; Sino Biological, Inc. Beijing, China) was diluted to 40 μg/mL and then immobilized on Ni–NTA biosensors (18–5114; Sartorius, NY, USA) for 300 s and an additional 60 s to obtain the baseline for verifying the protein-biosensor affinity. Calceolarioside B was serially diluted from 100 μM to 3.13 μM with PBS in a black plate (655997; Greiner Bio-One GmbH, Frickenhausen, Germany). After establishing the baseline for 120 s, the biosensors containing the RBD protein were immersed in wells with serial dilutions of calceolarioside B for 180 s to allow for association, followed by a dissociation step for another 180 s. The dissociation constant (KD) value was calculated by using a 1:1 binding model utilizing the Data Analysis Software 9.0 (Sartorius, USA).

Construction of the ACE2 receptor cell lines

Human embryonic kidney 293 T cell lines stably expressing ACE2-EGFP and ACE2-mcherry were constructed via lentiviral transfection. For this, ACE2-EGFP or ACE2-mcherry plasmids (Vectorbuilder Inc., Guangzhou, China) were used to co-transfect 293 T cells with the lentivirus packaging plasmids VSVG, GAG, and REV (#8454, #12251, #12253, Addgene Inc., MA, USA). After 12 h of incubation, the cells were supplied with a fresh medium. Following an additional 12 h, the supernatant containing the lentivirus was harvested and filtered through a 0.45 μm filter (SLHA033SB; Millipore, USA). Next, 5 × 104 293 T cells were seeded in 6-well plates and co-incubated with the lentivirus along with 5 μg/mL polybrene (Vectorbuilder Inc., China) for 48 h. After observing cell fluorescence, the cells were purified by treatment with 4 μg/mL puromycin (Cat: A1113803; Gibco, MT, USA) for 2 weeks.

Immunofluorescence of Omicron BA.2 protein binding to ACE2 293 T cells

1 × 104/well 293 T cells transfected with ACE2-EGFP plasmid were seeded on a coverslip placed inside a 24-well plate and incubated overnight. Spike-RBD protein (His Tag) (Cat: 40592-V08H123, Sino Biological, Inc.) and calceolarioside B (diluted from 200 to 25 μM) were premixed for 1 h and then incubated with the cells for another 1 h. The protein without premixing was the control. Subsequently, cells were washed and fixed with 1% paraformaldehyde (P804537-500 g; Shanghai Macklin Biochemical Co. Ltd., Shanghai, China) for 15 min, and blocked with 2% BSA (A801320-100 g; Shanghai Macklin) for 15 min. The cells were incubated with His Tag antibody (Cat: 12698S; Cell Signaling Technology, MA, USA) at room temperature for 2 h, followed by incubation with Alexa Fluor® 555 Phalloidin secondary antibody (Cat: 8953S; Cell Signaling Technology) for 2 h. After drying, the coverslips were mounted with the FluorSave™ reagent (345789; Calbiochem, Darmstadt, Germany). The binding of the S protein with the ACE2 receptors was visualized under a SP8 confocal microscope (Leica, Wetzlar, Germany). The semi-quantitative analysis was carried out using the ImageJ software (https://imagej.net/ij/download.html). The reduction in the density of secondary antibody-labeled Spike-RBD protein mediated by calceolarioside B was compared to the control group to assess its neutralizing potency.

Omicron BA.2 pseudovirus based virus neutralization assay

5 × 104/well 293T cells transfected with ACE2-mcherry plasmid were seeded overnight on a coverslip placed inside a 24-well plate. Calceolarioside B (diluted from 200 to 25 μM) was premixed with SARS-CoV-2 Omicron BA.2 spike pseudovirus and 5 μg/mL polybrene (Vectorbuilder Inc.) for 1 h to neutralize the virus. The mixtures were then added to the cells and incubated for 12 h. The medium was replaced and further incubated for 24 h. Next, the cells were washed and fixed with 1% PFA for 15 min and stained with DAPI (62248; Thermo-Fisher Scientific Inc., MA, USA) for 10 min. After the coverslips were dried, the slices were mounted using FluorSave reagent. The virus neutralization effect was visualized and analyzed employing the ImageXpress® Pico automated cell imaging system (Molecular Devices, CA, USA). The neutralizing potency was evaluated by calculating the percentage reduction of infected cells (marked in green).

MTT assay

Experiments were conducted using RLE-6TN rat type II alveolar epithelial cells (CRL-2300; ATCC, Manassas, VA, USA). To determine the half-maximal cytotoxic concentration (CC50) of calceolarioside B, cells were seeded into 96-well plates at a density of 5 × 10^3 cells per well. The cells were treated with calceolarioside B at concentrations of 0, 100, 200, 400, 800, 1600, and 3200 μM for 24 h, and cell viability was assessed using the MTT assay.

To determine the half-maximal effective concentration (EC50) of calceolarioside B, different concentrations of calceolarioside B (0, 12.5, 25, 50, 100, 200, and 400 μM) were mixed with pseudovirus and applied to RLE-6TN cells for 24 h. The EC50 was calculated based on the inhibition rate of viral infection. A selectivity index was calculated as the ratio CC50/IC50.

ELISA

To evaluate the effects of calceolarioside B on inflammation, RLE-6TN rat type II alveolar epithelial cells were induced with LPS (5 mg/L) to stimulate an inflammatory response. Subsequently, the cells were treated with calceolarioside B (200 μM) or left untreated. The Omicron BA.2 pseudovirus was added for stimulation. After 24 h, the cell culture supernatant was collected, and the IL-6 levels were measured using the Rat IL-6 ELISA Kit (E-EL-R0015c; Elabscience, Hubei, China) to determine the expression of the inflammation-related factors.

Flow cytometry

To observe the effect of calceolarioside B on the polarized state of macrophages, peripheral blood mononuclear cells (PBMCs) were purchased from Guangzhou Jennio Biotech Co. Ltd., Guangzhou, China. A transwell chamber composed of a 0.4 μm membrane was used to culture the monocytes in the lower chamber, and RLE-6TN rat type II alveolar epithelial cells in the upper chamber were then added with the medium (containing 200 μM calceolarioside) and Omicron BA.2 pseudovirus. After 72 h, the M1/M2 macrophages in the lower chambers of each group were stained with CD11b PE-Cy7 (60–0112; Tonbo Biosciences, CA, USA), CD163 (326510; BioLegend, CA, USA), CD206 (321104; BioLegend), and CD68 (333808; BioLegend) antibodies for 1 h. After incubation, the cells were washed with 2 mL PBS, then resuspended in 100 μL PBS, and detected by FACS Aria III flow cytometry (BD Inc., NJ, USA). Each experiment was repeated independently thrice.

Statistical analysis

Statistical analysis was conducted using the SPSS 22.0 software (IBM Corp., NY, USA), and the results were expressed as the mean ± standard deviation. The Shapiro–Wilk test was first used to judge whether the data conforms to normal distribution. All the data conformed to the normal distribution, one-way ANOVA was utilized to analyze the variations between multiple groups. If the results of the one-way ANOVA were significant, Tukey's method was utilized to perform the pairwise comparisons between the groups. A p value of < 0.05 was considered statistically significant.