*3.2. GA Effect on S Protein Binding to Host Cells*

Virus infection is initiated by receptor-mediated attachment, followed by cell entry via membrane fusion or endocytosis. Since Lenti-S is a pseudotyped virus that utilizes S protein for cell attachment and cell entry, we focused on whether GA targeted the S protein of SARS-CoV-2 for antiviral action. We performed a cell-binding assay using a biotin-labeled S protein to assess whether GA interfered with S protein attachment to host cells. In this regard, a biotinylated S protein was allowed to bind to Vero E6 cells in the presence or absence of GA (Figure 2A). After washing off non-bound S protein with an ice-cold medium, cell-bound S protein was detected by immunoblotting analysis. We used unlabeled S protein as a competitive reagent against biotin-labeled S protein binding to demonstrate the binding specificity of biotinylated S protein to host cells (Figure 2B). As shown in Figure 2C, GA treatment dose-dependently inhibited S protein binding to the cells. At 5 mM concentration, GA blocked S protein binding by more than 95%, suggesting that GA inhibited pseudovirus infection likely by blocking S protein-mediated cell attachment.

**Figure 2.** Effect of GA on recombinant S protein binding. (**A**) Schematic presentation of the experimental design. The cells were detached and incubated on ice for 60 min with GA or recombinant S protein (S unlabeled). Biotinylated S protein was then added for cell-attachment assay. After washing with ice-cold DMEM, cell-attached biotinylated S protein was detected by immunoblotting assay. (**B**) Biotinylated S protein binding to host cells in the presence of BSA or unlabeled S protein. Vero E6 cells resuspended in ice-cold DMEM containing 0.5% BSA were allowed to interact with Lenti-S in the presence or absence of unlabeled S protein (at 2 and 10 μg/mL). (**C**) Effect of GA on S protein binding to Vero E6 cells. A biotinylated S protein was allowed to bind to detached Vero E6 cells in the absence or presence of GA. Cell-bound biotinylated S protein was determined by immunoblotting assay. Actin was used as a loading control. Con, Vero E6 cells only. Input, biotinylated S protein for total binding.

#### *3.3. GA Treatment of Pseudovirus but Not the Cells Inhibits Pseudovirus Infection*

We performed an infection assay by pretreatment of the virus and the cells to preliminarily determine a primary target of the GA effect. In this regard, Lenti-S was pretreated with GA at a concentration of 3 mM on ice for 1 h. The treated virus was then 1:30 diluted with culture medium and used to infect the cells (final GA concentration in the culture medium was at approximately 0.1 mM; the concentration showed no antiviral activity). Virus infection was determined by measuring luciferase activity at 24 h post infection. Alternatively, we also treated the cells with 3 mM GA on ice for 1 h to determine whether GA targeted the host cells for its antiviral effect since the receptor ACE2 was a putative target of GA action [22]. After removal of GA by rinsing with fresh DMEM, the cells were then infected with Lenti-S for 24 h. Pretreatment of cells with GA showed marginal

effect against Lenti-S infection, while pretreatment of pseudovirus with GA profoundly reduced pseudovirus infectivity (Figure 3), indicating that GA targeted virus particles for the antiviral effect.

**Figure 3.** Pretreatment of Lenti-S or host cells with GA on Lenti-S mediated luciferase gene delivery. (**A**) Diagram showing Lenti-S or host cells were treated (Tx) with 3 mM GA prior to the infection assay. Both Lenti-S and host cells were untreated or treated with 3 mM GA for 1 h. At the end, the GA-treated Lenti-S was 1:30 diluted and used to infect untreated host cells (final concentration of GA in the medium was approximately 0.1 mM). In parallel, the medium of the GA-treated cells was replaced with fresh medium without GA followed by infection of untreated Lenti-S. Luciferase activity was determined 24 h later. (**B**) GA effect on Lenti-S (Virus Tx) and on host cells (Cells Tx). Luciferase activity was expressed as a percentage of untreated controls. Data are mean ± SD of duplicate wells from 2 experiments. ns, no significance; \*\*, *p* < 0.01 by Student's *t* test.

#### *3.4. GA Interacts with S Protein*

To more clearly demonstrate whether GA interacted with S protein, we performed an SPR assay to measure GA interaction with a recombinant S protein (Figure 4). The *k*a and *<sup>k</sup>*<sup>d</sup> were measured at approximately 7.6 × <sup>10</sup><sup>5</sup> <sup>M</sup>−1s−<sup>1</sup> and 2.2 × <sup>10</sup>−<sup>4</sup> <sup>s</sup>−1, respectively, which translated to a calculated *K*<sup>D</sup> of 0.28 nM, if a 1:1 stoichiometry was used for GA and S protein interaction. The result shows clearly a direct interaction between GA and S protein, although this might have overestimated the affinity since the S protein is believed to be a homotrimeric protein which would have more than one GA binding site per protein. Regardless, the results together indicate that GA blocked Lenti-S infection through inhibition of S protein-mediated cell binding.

**Figure 4.** Determination of GA binding to S protein by SPR. S protein was immobilized on to a CM5 sensor chip. GA at concentrations as indicated was passed over the chip, and SPR angle changes were recorded using the Biacore T200 system and reported as response units (RUs). Data fitting was performed using the 1:1 Langmuir model in the BIAevaluation software package (GE Healthcare).

### *3.5. Autodocking Reveals GA-Binding Pockets on SARS-CoV-2 S Protein*

It was predicted that the RBD of the S protein contains binding pockets for natural products including GA [23]. We performed AutoDock Vina analysis by scanning through the entire extracellular domain of the S protein for the binding potentials. The first notion was GA might bind on the interaction interface of S protein–ACE2 since we found that GA treatment blocked Lenti-S infection and S protein attachment. Indeed, a binding pocket at the S-ACE2 interface was identified with a calculated binding energy of −8.0 kcal/mol (Figure 5A). The S protein presents in two different conformations, named open and closed states [24]. We also identified another binding pocket located at the inner side of the RBD with a binding energy of −7.0 kcal/mol (Figure 5B). SARS-CoV-2 opening is expected to be necessary for interacting with ACE2 at the host-cell surface and initiating the conformational changes leading to cleavage of the S2 site for efficient membrane fusion and viral entry [25]. It is possible that GA binding at this inner binding pocket impacted the close–open state transformation resulting in diminished open-state trimer to interact with ACE2. Thus, we also searched for the potential binding pocket at the ACE2 receptor. Similar to a previous report [22], a pocket at the interface binding to the S protein provided a binding energy of −4.1 kcal/mol (Figure 5C). Based on the predicted binding energy, we speculated that the primary binding site of GA should be on the S protein rather than the ACE2.

**Figure 5.** Autodocking analysis of GA interaction with SARS-CoV-2 S and ACE2 protein. Three protomers of SARS-Cov-2 S protein (PDB id: 6vsb) are shown in cyan, green, and yellow, respectively. ACE2 protein (PDB id: 6m18) is in orange. The structure of proteins is presented in ribbons. The structure of GA (ZINC id:960251743495) is shown in magenta as sticks. Arrows indicate the predicted binding site of GA. (**A**) Predicted binding of GA on the S protein at the S–ACE2 interface with a calculated binding energy of –8.0 kcal/mol. (**B**) Predicted binding of GA on a binding pocket located at the inner side of the RBD with a binding energy of −7.0 kcal/mol. (**C**) Predicted binding of GA on the ACE2 protein at the ACE2–S interface with a binding energy of −4.1 kcal/mol.

In summary, we showed here that GA interacted with SARS-CoV-2 S protein and blocked S protein-mediated cell binding for the antiviral activity of GA against SARS-CoV-2.

#### **4. Discussion**

Coronaviruses are a group of related viruses that cause diseases in humans and animals. In humans, coronaviruses cause respiratory tract infections, ranging from the common cold to the deadly diseases by SARS-CoV, MERS-CoV, and SARS-CoV-2. Due to the lack of medicines for COVID-19, repurposing currently existing and experimental drugs has been proposed as an alternative to uncover agents with therapeutic potentials. Traditional medicines have demonstrated records as anti-infectives throughout the history of mankind and have shown to be effective in China at alleviating COVID-19 symptoms

or even reducing fatality. GA, a major component of *Glycyrrhiza* spp., possesses a wide range of pharmacological and biological activities, including antioxidant, antiviral, and anti-inflammatory effects [7,13,26,27]. Gowda and colleagues reported that GA could inhibit SARS-CoV-2-protein-induced high-mobility group box 1 (HMGB1) release and inhibits viral replication [28]. GA also targets SARS-CoV-2 main protease [29] and blocks proinflammatory response [30]. It is likely that GA can utilize multiple mechanisms against SARS-CoV-2 infection and disease [29,30]. To initiate a productive cycle of infection, a virus first attaches to a host cell, followed by a cell entry and replication process. In this study, we used a pseudotyped lentivirus and showed GA with antiviral activity. We focused on the early stages of virus infection using a pseudotyped virus system. This approach has been successfully used to construct pseudotyped lentiviruses for SARS-CoV, MERS-CoV, and recently SARS-CoV-2 and the corresponding mutants [31–34]. As a model for highly contagious pathogens, pseudotyped viruses are easy to construct and safe to use. It allows in particular the detailed studies involving virus attachment and virus-cell entry stages.

Several studies have predicted S protein or S–ACE2 interaction as potential targets for GA against SARS-CoV-2 [27,35]. Here we provided experimental evidence demonstrating that GA blocks S protein-mediated cell attachment for its antiviral effect. We found that GA interacted with the S protein with high affinity and blocked a recombinant S protein binding to the host cells. We also executed computational molecular docking to elucidate potential GA binding pockets on S protein. We screened through the entire extracellular domain of S protein by defining multiple grid boxes within this region. In addition to a previously revealed GA binding site at the interaction interface between the RBD and ACE2 protein [36], we also found a binding pocket at the inner side of the RBD. Based on the structural features, we predicted that the binding may have several impacts on the infectious activity of Lenti-S. First, S protein presents in two different conformations including a close state and an open state with one RBD of the trimer flipped out. The switch from close state to open state of S protein was necessary to establish an interaction with the ACE2 receptor. The binding site at the inner face of the RBD could impact this conformation transition. Secondly, it is also likely that GA binding at the inner side of the RBD might interfere with subsequent conformational change during the fusion stage.

In an assimilated SARS-CoV-2-infected mouse model, nanoparticles carrying GA demonstrated therapeutic effects through anti-inflammatory and antioxidant activities [37]. At the intracellular and circulating levels, GA binds to high-mobility group box 1 protein (HMGB1) to provide robust anti-inflammatory and neuroprotection [38,39]. GA attenuates pulmonary hypertension progression and pulmonary vascular remodeling in animal models [40–42]. As a hydrophilic compound, GA is not readily absorbed. After oral ingestion, glycyrrhizin is first hydrolyzed to 18 β-glycyrrhetinic acid by intestinal bacteria, which can be absorbed from the gut [43]. The metabolites in circulation, along with GA, can significantly reduce inflammatory cell infiltration and cytokine production during an infection [40,42].

Pompei and colleagues reported that GA was effective against a broad range of enveloped viruses [10]. Cinatl et al. showed that GA was effective against SARS-CoV [12,44]. The reported concentrations for GA antiviral effect in cell cultures generally vary between 1 and 5 mM concentrations, at which concentrations GA can form an emulsion or long-lasting foams in an aqueous solution [45–47]. It is well known that surfactants were able to inactivate enveloped viruses by causing protein aggregation, disruption of the envelope, or by distorting the shape of virions [48]. At the membrane level, GA induces cholesterol-dependent disorganization of lipid rafts which are important for the entry of coronavirus into cells [49,50]. GA was also reported to modulate the fluidity of the plasma membrane and HIV-1 envelope [45]. The fact that GA directly inactivated enveloped virus particles suggests that GA likely exerts its antiviral activity by destabilizing the envelope. Whether chemicals such as GA use the surfactant activity for their antiviral effect remains to be further studied.

Here we provided experimental and computational simulation data demonstrating that GA potentially targets S protein-mediated cell attachment for its antiviral activity. GA interacted with the S protein with high affinity and blocked recombinant S protein binding to the host cells. Thus, this study uncovered a mechanism by which GA blocks SARS-CoV-2 infection, highlighting the potential of herbal medicine against emerging and reemerging infectious diseases.

**Author Contributions:** S.H., E.L., and J.L. conceived the concepts. J.L., D.X., L.W., S.H., G.Z., and M.Z. performed the experiments and analyzed the data. E.L., S.H., and J.L. curated the data. E.L., L.W., and S.H. wrote the draft. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by grants from Ningxia Hui Autonomous Region (2017BN04, E.L.), Jiangsu Natural Science Foundation (BK20200316 to S.H.), NSFC (81871636 to E.L.), Central Universities Fundamental Research Funds (14380456 to E.L. and 14380470 to S.H.), Science, Technology and Innovation Commission of Shenzhen Municipality (JSGG 20200519160755008 to E.L.), and Nanjing Scientific and Technological Innovation Project for Returning Scholars from Abroad: (2021 to S.H.).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** We thank Guang Yang for technical assistance and Jinhui Dou for comments.

**Conflicts of Interest:** The authors declare no conflict of interest.

**Sample Availability:** Not available.

#### **Abbreviations**


#### **References**


MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Molecules* Editorial Office E-mail: molecules@mdpi.com www.mdpi.com/journal/molecules

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com

ISBN 978-3-0365-2713-0