New Phenolic Lipids from the Leaves of Clausena harmandiana Inhibit SARS-CoV-2 Entry into Host Cells

Induced by the spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the COVID-19 pandemic underlined the clear need for antivirals against coronaviruses. In an effort to identify new inhibitors of SARS-CoV-2, a screening of 824 extracts prepared from various parts of 400 plant species belonging to the Rutaceae and Annonaceae families was conducted using a cell-based HCoV-229E inhibition assay. Due to its significant activity, the ethyl acetate extract of the leaves of Clausena harmandiana was selected for further chemical and biological investigations. Mass spectrometry-guided fractionation afforded three undescribed phenolic lipids (1–3), whose structures were determined via spectroscopic analysis. The absolute configurations of 1 and 2 were determined by analyzing Mosher ester derivatives. The antiviral activity against SARS-CoV-2 was subsequently shown, with IC50 values of 0.20 and 0.05 µM for 2 and 3, respectively. The mechanism of action was further assessed, showing that both 2 and 3 are inhibitors of coronavirus entry by acting directly on the viral particle. Phenolic lipids from Clausena harmandiana might be a source of new antiviral agents against human coronaviruses.


Introduction
In late 2019, a new coronavirus named SARS-CoV-2 (Severe Acute Respiratory Syndrome CoronaVirus 2) emerged in Wuhan, China, causing a global pandemic. Those with COVID-19 present symptoms of viral pneumonia, and the virus can cause fatal respiratory illness [1]. This novel coronavirus disease spread rapidly around the world, affecting tens of millions of people. Although vaccines have been developed, safe and efficient antiviral treatments are still needed to control the emergence of such viruses.
Recently, with the aim of discovering new antiviral agents, we have developed a molecular networking-based strategy that revolves around deciphering the relationship between spectral networks and biological activities and further exploiting such a relationship to prioritize the isolation of bioactive natural products [2][3][4]. In the present study, this approach allowed us to target the isolation of specific compounds from the ethyl acetate

Results and Discussion
Within a project aiming to investigate the potential antiviral activity of medicinal plants belonging to the Rutaceae and Annonaceae families, a total of 824 EtOAc plant extracts previously filtered on polyamide cartridge were screened for the inhibition of Human alphacoronavirus (HCoV-229E) as a coronavirus model. Huh-7 cells and Huh-7-TMPRSS2 cells were infected with a recombinant molecular clone expressing the luciferase gene reporter HCoV-229E-Luc, which allowed for rapid screening of the several hundred plant extracts. The collection of Rutaceae evaluated included eleven extracts from different plant parts of three species of the genus Clausena. Of these, only the crude EtOAc extract of the leaves of C. harmandiana showed potent antiviral activity; therefore, the crude EtOAc extract was selected for further chemical and biological investigations.
All Clausena extracts were profiled by LC-HRESIMS 2 as previously described [2][3][4]. Briefly, the samples were analyzed using a data-dependent acquisition mode. The resulting spectral data were preprocessed via MZmine 2 [10] and structured into molecular networks [11] using Metgem [12] ( Figure 1A). The color code indicates which extract(s) each ion comes from. A cluster quasi-specific to the leaf extract of C. harmandiana ( Figure 1B, in red) was highlighted. As the only active extract, the compounds corresponding to these ions were most likely responsible for the antiviral activity. Therefore, they were subjected to targeted isolation. The EtOAc crude extract was first subjected to silica gel flash column chromatography, yielding 11 fractions (F1-F11); subsequently, the fractions were analyzed using LC-HRESIMS. F4 were shown to contain compounds corresponding to the cluster of interest and, with an IC 50 value of 0.06 µg/mL, displayed strong anti-HCoV-229E activity. From this fraction, C-18 column chromatography afforded the three undescribed phenolic lipids (referred to here as phenolic lipids 1-3, the structures of which are shown in Figure 2).

Structural Elucidation
Compound 1 was obtained as a gray amorphous powder.      The positions of the three hydroxy groups at C-1, C-4, and C-2 and the double bond at C-14 /C-15 were confirmed by 2D NMR analyses ( Figure 3). The para position of phenol groups was confirmed from the HMBC correlations from H-1 (∂ H 2.75) to C-3 (∂ C 118.4); and H-3 (∂ H 6.77) to C-1 (∂ C 39.2). In addition, the chemical shift values of the aromatic carbons are similar to those reported in the literature for close analogues [13]. The stereochemistry at the double bond between C-14 and C-15 was assigned as Z on the basis of allylic carbon resonances at ∂ C 27.6 and 29.6 ppm (C-13 and C-16 ) [14] and to similar values of coupling constants of structurally close alkenylresorcinols [15].
Due to the specific rotation value of 0 for compound 1, we initially assumed that this compound could be a racemic mixture. However, it has been reported in the literature that enantiomerically pure [16]. Thus, to establish the absolute configuration of compound 1, Mosher ester derivatization experiments were performed [17]. The secondary alcohol was converted into the (S)-and (R)-MPTA esters (1a and 1b) following the procedure described in Brel et al. [18]. Based on the ∆δ (δ S-δ R ) values of both MPTA esters (Figure 4), the R absolute configuration of C-2 was determined.
Due to the specific rotation value of 0 for compound 1, we initially assumed that th compound could be a racemic mixture. However, it has been reported in the literatu that choerosponols B and C, which possess similar planar structures and D values of are enantiomerically pure [16]. Thus, to establish the absolute configuration of compoun 1, Mosher ester derivatization experiments were performed [17]. The secondary alcoh was converted into the (S)-and (R)-MPTA esters (1a and 1b) following the procedure d scribed in Brel et al. [18]. Based on the Δδ (δS-δR) values of both MPTA esters (Figure 4), th R absolute configuration of C-2′ was determined.  . NMR data were almost th same to those of 1 but suggested the presence of one additional methylene. Indeed, th 13 C-NMR spectrum (Table 1)  The NMR data of compound 3 were almost identic to those of 1 but without a double bond, as deduced from the absence of carbon aroun 130 ppm in the 13 C-NMR spectrum (Table 1). Based on biogenetic considerations, the ( absolute configuration of C-2′ was proposed, as for compounds 1 and 2.

Cytotoxicity and H-CoV-229E Inhibition Assays
Compounds 1-3 were first evaluated for their cytotoxic activities against Huh7, a h man hepatocyte cell line. All compounds showed cytotoxic activities in the micromol range (CC50 between 0.5 and 1.3 µM) ( Table 2). The maximum non-toxic concentration th provides more than 95% viability was 0.25 µM for all three tested compounds. This co centration was then used as the highest concentration for antiviral assays. Thus, to asse . NMR data were almost the same to those of 1 but suggested the presence of one additional methylene. Indeed, the 13 C-NMR spectrum (Table 1)  The NMR data of compound 3 were almost identical to those of 1 but without a double bond, as deduced from the absence of carbon around 130 ppm in the 13 C-NMR spectrum (Table 1). Based on biogenetic considerations, the (R) absolute configuration of C-2 was proposed, as for compounds 1 and 2.

Cytotoxicity and H-CoV-229E Inhibition Assays
Compounds 1-3 were first evaluated for their cytotoxic activities against Huh7, a human hepatocyte cell line. All compounds showed cytotoxic activities in the micromolar range (CC 50 between 0.5 and 1.3 µM) ( Table 2). The maximum non-toxic concentration that provides more than 95% viability was 0.25 µM for all three tested compounds. This concentration was then used as the highest concentration for antiviral assays. Thus, to assess whether compounds 1-3 exert antiviral activity against HCoV-229E, Huh7 cells were infected with the recombinant molecular clone of HCoV-229E that expresses the luciferase (HCoV-229E-Luc) at the MOI of 0.5 in the presence of different concentrations (serial dilution starting from 0.25 µM) of each compound throughout the infection [19]. The results indicated that compound 1 showed no antiviral effect at non-cytotoxic doses, while compounds 2 and 3 exert dose-dependent antiviral activity at non-cytotoxic concentrations, with IC 50 values of 0.1 and 0.05 µM, respectively, resulting in a selectivity index (SI) of 5 and 16 for 2 and 3, respectively (Table 2). 0.50 ± 0.05 0.10 ± 0.03 5 3 0.80 ± 0.10 0.05 ± 0.04 16 Cytotoxic concentration (CC 50 ) and inhibitory concentration (IC 50 ) were obtained by performing nonlinear regression followed by the construction of the sigmoidal concentration-response curves from Figure 4. a Concentration inhibited cell viability by 50%; b Concentration inhibited infection by 50%; c Selectivity index (CC 50 /IC 50 ). na: non active.

Cytotoxicity and SARS-CoV-2 Inhibition Assays
Prior to the assessment of their antiviral activity against SARS-CoV-2, the cytotoxicity of compounds 1-3 was determined against Vero-E6 cells ( Figure 5A). MTS assays showed that all tested compounds exert similar dose-dependent cytotoxicity on Vero-E6 cells with 50% cytotoxic concentration (CC 50 ) values of 1.5 ± 0.20, 0.5 ± 0.08, and 0.9 ± 0.20 µM for 1, 2, and 3, respectively ( Figure 5A). Then, the antiviral activity of the three compounds was evaluated against SARS-CoV-2 in Vero-E6 cells. For this, Vero cells were infected with the pandemic strain of SARS-CoV-2 at an MOI of 0.1 for 48 h in the presence of different non-cytotoxic concentrations of the compounds ( Figure 5B). The results showed that only 2 and 3 exert dose-dependent antiviral activity against SARS-CoV-2 at non cytotoxic concentrations ( Figure 4B), with IC 50 values of 0.20 ± 0.06 and 0.05 ± 0.03 µM, respectively. The selectivity index (SI) values-2.5 and 18 for 2 and 3, respectively-are similar to those obtained with HCoV-229E. Taken together, these results showed that 3 is a strong inhibitor of SARS-CoV-2 infection.
indicated that compound 1 showed no antiviral effect at non-cytotoxic doses, while compounds 2 and 3 exert dose-dependent antiviral activity at non-cytotoxic concentrations, with IC50 values of 0.1 and 0.05 µM, respectively, resulting in a selectivity index (SI) of 5 and 16 for 2 and 3, respectively ( Table 2). Cytotoxic concentration (CC50) and inhibitory concentration (IC50) were obtained by performing nonlinear regression followed by the construction of the sigmoidal concentration-response curves from Figure 4. a Concentration inhibited cell viability by 50%; b Concentration inhibited infection by 50%; c Selectivity index (CC50/IC50). na: non active.

Cytotoxicity and SARS-CoV-2 Inhibition Assays
Prior to the assessment of their antiviral activity against SARS-CoV-2, the cytotoxicity of compounds 1-3 was determined against Vero-E6 cells ( Figure 5A). MTS assays showed that all tested compounds exert similar dose-dependent cytotoxicity on Vero-E6 cells with 50% cytotoxic concentration (CC50) values of 1.5 ± 0.20, 0.5 ± 0.08, and 0.9 ± 0.20 µM for 1, 2, and 3, respectively ( Figure 5A). Then, the antiviral activity of the three compounds was evaluated against SARS-CoV-2 in Vero-E6 cells. For this, Vero cells were infected with the pandemic strain of SARS-CoV-2 at an MOI of 0.1 for 48 h in the presence of different noncytotoxic concentrations of the compounds ( Figure 5B). The results showed that only 2 and 3 exert dose-dependent antiviral activity against SARS-CoV-2 at non cytotoxic concentrations ( Figure 4B), with IC50 values of 0.20 ± 0.06 and 0.05 ± 0.03 µM, respectively. The selectivity index (SI) values-2.5 and 18 for 2 and 3, respectively-are similar to those obtained with HCoV-229E. Taken together, these results showed that 3 is a strong inhibitor of SARS-CoV-2 infection.

Characterization of the Antiviral Mechanism of Action of Compounds 2 and 3
To gain insights into the mechanism of action of 2 and 3 against coronaviruses, different experimental approaches were performed during HCoV-229E infection. Compounds were added at different stages of the viral replication cycle. To assess the impact on viral entry stage, HCoV-229E and compounds 2 and 3 were simultaneously co-added to the cells for 1 h ( Figure 6A). Isoquercitrin (Q3G), which is known to inhibit the endocytic pathway, was used as a positive control [20], and compound 1 served as a negative control. To investigate whether 2 or 3 interfere with HCoV-229E replication, cells were first challenged with HCoV-229E for 1 h and then treated with 2 or 3 ( Figure 6A). Remdesivir, which is known to inhibit virus replication, was used as a positive control [21].
To gain insights into the mechanism of action of 2 and 3 against coronaviruses, different experimental approaches were performed during HCoV-229E infection. Compounds were added at different stages of the viral replication cycle. To assess the impact on viral entry stage, HCoV-229E and compounds 2 and 3 were simultaneously co-added to the cells for 1 h ( Figure 6A). Isoquercitrin (Q3G), which is known to inhibit the endocytic pathway, was used as a positive control [20], and compound 1 served as a negative control. To investigate whether 2 or 3 interfere with HCoV-229E replication, cells were first challenged with HCoV-229E for 1 h and then treated with 2 or 3 ( Figure 6A). Remdesivir, which is known to inhibit virus replication, was used as a positive control [21]. Our data showed that no inhibition of infection was observed when 2 or 3 were added after virus inoculation, suggesting that 2 and 3 do not affect the virus replication step, unlike remdesivir (positive control), which strongly inhibited viral replication ( Figure 6B). In contrast, strong inhibition of infection was noticed when 2 and 3 were present during the virus entry step ( Figure 6B), as well as the positive control Q3G. Taken together, these results suggest that 2 and 3 act as virus entry inhibitors.
To further elucidate the underlying mechanism of antiviral action, we investigated whether 2 and 3 target the virus or the cells. HCoV-229E particles were pre-incubated with 2 or 3 (0.2 µM) for 1 h and then diluted 20-fold prior to infection to reach a concentration  Our data showed that no inhibition of infection was observed when 2 or 3 were added after virus inoculation, suggesting that 2 and 3 do not affect the virus replication step, unlike remdesivir (positive control), which strongly inhibited viral replication ( Figure 6B). In contrast, strong inhibition of infection was noticed when 2 and 3 were present during the virus entry step ( Figure 6B), as well as the positive control Q3G. Taken together, these results suggest that 2 and 3 act as virus entry inhibitors.
To further elucidate the underlying mechanism of antiviral action, we investigated whether 2 and 3 target the virus or the cells. HCoV-229E particles were pre-incubated with 2 or 3 (0.2 µM) for 1 h and then diluted 20-fold prior to infection to reach a concentration of 0.01 µM for inoculation ( Figure 7A), a concentration that does not inhibit HCoV-229E-Luc infection for both 2 and 3 (as shown above). Compound 1 was used as a negative control. In parallel, Huh7 cells were infected with HCoV-229E-Luc and subsequently treated with 0.2 and 0.01 µM of 2 or 3, serving as controls ( Figure 7B). Phospholipase (PLA2), which is known to have broad-spectrum virucidal activity, was used as a positive control [22]. The results clearly indicated that when HCoV-229E-Luc was pre-incubated with 2 or 3 at a high concentration (0.2 µM) before infection at a low concentration (0.01 µM), the antiviral activity was much stronger than when infection was performed in the presence of 0.01 µM of 2 or 3 without pre-incubation ( Figure 7B). Compounds 2 and 3 were able to inhibit the infection up to 60 and 80%, respectively, as well as PLA2, which inhibited 90% of infection ( Figure 7B). Taken together, these results suggest that 2 and 3 inhibit HCoV229E entry by acting directly on the viral particle. of 0.01 µM for inoculation ( Figure 7A), a concentration that does not inhibit HCoV-229E-Luc infection for both 2 and 3 (as shown above). Compound 1 was used as a negative control. In parallel, Huh7 cells were infected with HCoV-229E-Luc and subsequently treated with 0.2 and 0.01 µM of 2 or 3, serving as controls ( Figure 7B). Phospholipase (PLA2), which is known to have broad-spectrum virucidal activity, was used as a positive control [22]. The results clearly indicated that when HCoV-229E-Luc was pre-incubated with 2 or 3 at a high concentration (0.2 µM) before infection at a low concentration (0.01 µM), the antiviral activity was much stronger than when infection was performed in the presence of 0.01 µM of 2 or 3 without pre-incubation ( Figure 7B). Compounds 2 and 3 were able to inhibit the infection up to 60 and 80%, respectively, as well as PLA2, which inhibited 90% of infection ( Figure 7B). Taken together, these results suggest that 2 and 3 inhibit HCoV229E entry by acting directly on the viral particle.    To assess whether the antiviral activity of 2 and 3 against SARS-CoV-2 is also attributable to their ability to inhibit virus infectivity by acting directly on the virus particle, a residual infectivity assay was carried out. SARS-CoV-2 virus particles were incubated with 2 or 3 (0.2 µM). Compound 1 (0.2 µM) and PLA2 (5 µg/mL) were used as negative and positive controls, respectively. After 1 h of incubation, titration of residual infectivity showed that compounds 2 and 3 were able to decrease viral progeny up to 1 log as well as the positive control. Compound 3 appears to be the most effective at a non-cytotoxic dose of 0.2 µM, where the molecule is able to significantly inhibit SARS-CoV-2 infectivity by more than 1 log. These results suggest that compounds 2 and 3 possess antiviral activity against SARS-CoV-2 by acting directly on the virus particle.
From a structural standpoint, compounds 1-3 are monosubstituted hydroquinone and belong to a large family of phenolic lipids. They are structurally very similar to choerosponols B and C isolated from Choerospondias axillaris, an Anacardiaceae native to Nepal [16]. They all possess a C-2 hydroxylated unsaturated (1 and 2) or saturated (3) alkyl side chain, which only differs in the number of carbons: 18 for 1 and 3; 19 for 2. As choerosponols and with IC 50 values in the micromolar range, compounds 1-3 exhibited strong cytotoxic activities on different cell lines. In contrast, only compounds 2 and 3 exert strong antiviral activities against HCoV-229E and SARS-CoV-2 viruses; compound 1 shows no antiviral activity. While the presence of a double bond in the C-14 position of the alkyl side chain of compound 1 could explain the activity difference between compounds 1 and 3, it cannot explain the difference regarding compound 2, which only possesses an additional methylene group on the alkyl chain. Therefore, the discrepancy between the antiviral activity of 1 and 2 is difficult to explain rationally, and further studies are needed to explain it better.
Multi-informative molecular networks, which compare taxonomically related samples, are a highly useful approach for highlighting a chemical family of interest within a bioactive extract. In the present study, we used molecular networks in conjunction with biological data, which enabled us to directly target the isolation of two compounds with strong antiviral activities.

Plant Material
The leaves of Clausena harmandiana were collected in Hòa Bình, Mai Châu, Vietnam, in April 1996 and identified by Dr. Nguyen Cuong. A voucher specimen (VN-0081) was deposited at the Institute of Ecology and Biological Resources, Vietnam Academy of Science and Technology (VAST), Hanoï.

Extraction and Isolation
Dried leaves of C. harmandiana (150 g) were extracted with EtOAc (3 × 200 mL, 1 h each at room temperature). The EtOAc solutions were combined and evaporated to dryness under reduced pressure to give a crude residue (1.2 g). This residue was subjected to flash chromatography over silica gel and eluted with a gradient of Heptane-EtOAc (90:10 to 0:100) then EtOAc-MeOH ( Table 1.

Preparation of (S)-MTPA and (R)-MTPA Esters of 1 and 2
In an NMR sample tube, 1.0 mg of dimethylaminopyridine (DMAP) was added to 1.0 mg of compound 1 or 2. Additionally, 9 µL of pyridine-d 5

Data-Dependent LC-ESI-HRMS 2 Analysis
LC analyses were performed using a Thermo Ultimate 3000 system equipped with a Cortecs C 18 column (2.1 × 100 mm; 2.7 µm, Waters). The mobile phase consisted of water-acetonitrile (H 2 O-CH 3 CN) acidified with 0.1% formic acid (90:10) held for 2 min, then a gradient from 90:10 to 0:100 in 20 min held at 0:100 for 8 min at a flow rate of 600 µL.min −1 . The temperature of the column oven was set to 40 • C, and the injection volume was set to 5 µL. LC-ESI-HRMS 2 analyses were achieved by coupling the LC system to an Impact II Bruker quadrupole time-of-flight mass spectrometer (Bruker Daltonics, Bremen, Germany) equipped with an ESI dual source (operating in the positive-ion mode). Source parameters were set as follows: end plate offset-350 V, capillary voltage-4500 V, nebulizer pressure-60 psi, drying gas flow rate-10 L.min −1 , drying gas temperature-240 • C. MS scans were operated in full-scan mode from m/z 100 to 1400 (at 6 Hz). MS 1 scan was followed by MS 2 scans of the five most intense ions above an absolute threshold of 2000 counts. Selected parent ions were fragmented with a collision energy fixed at 30 eV and using an m/z dependent isolation window of 2-4 amu. The mass accuracy was guaranteed via an injection of a calibration solution from sodium formate clusters with external (at the beginning of each run) and internal (segment 0.1 at 0.4 min of each sample) calibration by a High Precision Calibration (HPC) equation with a maximum mass delta of 1 ppm and 7 as the minimal number of calibration points. LC-UV and MS data acquisition and processing were performed using DataAnalysis 4.4 software (Bruker Daltonics, Bremen, Germany).

MZmine 2 Pre-Processing
The MS 2 data files were converted from the .d Agilent standard data format to the .mzXML format using MSConvert software (part of the ProteoWizard package (Palo Alto, CA, USA, v3)) [23]. The .d Bruker data files were converted to .mzXML format using DataAnalysis 4.4 software. All .mzXML were then processed using MZmine 2 v53 [10]. Mass detection was conducted using a noise level of 400 counts for MS and 0 count for MSMS dimension. The ADAP chromatogram builder was used with a minimum group size of scans of 4, a group intensity threshold of 3000, a minimum highest intensity of 4000, and an m/z tolerance of 15 ppm [24]. The ADAP wavelets deconvolution algorithm was used with the following standard settings: S/N threshold = 8, minimum feature height = 3000, coefficient/area threshold = 10, peak duration range-0.02-1.0 min, RT wavelet range-0.01-0.07. Isotopologues were grouped using the isotopic peaks grouper algorithm, with an m/z tolerance of 15 ppm and an RT tolerance of 0.1 min. MS 2 scans were paired using an m/z tolerance range of 0.025 Da and RT tolerance range of 0.1 min. Peak alignment was performed using the join aligner module (m/z tolerance = 15 ppm, weight for m/z = 1, weight for RT = 1, absolute RT tolerance = 0.1 min). The peak list was gap-filled with the peak finder module (m/z tolerance = 5 ppm and RT tolerance = 0.05 min). Eventually, the .mgf spectral data file and its corresponding .csv metadata file (containing RT and peak areas) were exported using the dedicated "Export to GNPS-FBMN" built-in module [11,25].

Molecular Network Analysis
The two files mentioned above were imported into MetGem 1.3.6. [12]. MS 2 spectra were window-filtered by choosing only the top ten peaks within the ±50 Da window throughout the spectrum. The data were filtered by removing all peaks in the ±5 Da range around the precursor m/z. The m/z tolerance window used to find the matching peaks was set to 0.02 Da, and cosine scores were kept under consideration for spectra sharing at least 4 matching peaks. The network was created where edges were filtered to have a cosine score above 0.8. Further edges between two nodes were kept in the network only if each of the nodes appeared in each other's respective top 10 most similar nodes.

Cytotoxic Assays
The cytotoxicity of isolated phytocompounds were determined using an MTS [3-(4,5dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium]-based viability assay (CellTiter Aqueous One Solution Cell Proliferation Assay from Promega, Charbonnières-les-Bains, France). A total of 2 × 10 4 Huh7 and Vero-E6 cells were seeded on a 96-well plate and incubated with a serial dilution of phytocompounds. Forty-eight hours after treatment, the MTS test was performed according to the manufacturer's instructions. Absorbance was assessed at 490 nm, and the percentage of viable cells was calculated. Doseresponse curves were established on Prism GraphPad to calculate the concentration causing death in 50% of the cells (CC 50 ).

HCoV-229E-Luc Infection Inhibition Assays (Screening of 824 Plant Extracts)
HCoV-229E-Luc was mixed with the plant extracts at three different concentrations (25, 10 and 2.5 µg/mL) for 10 min. Huh-7 cells and Huh-7 cells transduced with a lentiviral vector expressing the TMPRSS2 protease gene (Huh-7-TMPRSS2 cells) were infected with HCoV-229E-Luc at a multiplicity of infection (MOI) of 0.5 in a final volume of 50 µL for 1 h at 37 • C in the presence of the plant extracts. The virus was removed and replaced with culture medium containing the extracts for 6 h at 37 • C. Cells were lysed in 20 µL of Renilla lysis buffer (Promega), and luciferase activity was quantified in a Tristar LB 941 luminometer (Berthold Technologies, Bad Wildbad, Germany) using a Renilla luciferase assay system (Promega), as recommended by the manufacturer.

HCoV-229E-Luc Infection Inhibition Assays (Evaluation of Fractions and Pure Compounds)
Huh7 cells were treated with different concentrations of fractions or compounds diluted in the culture medium and inoculated with HCoV-229E at a MOI of 0.5 in a final volume of 100 µL. At 12 h post-infection, the medium was removed, and the cells were lysed in 40 µL of Renilla luciferase buffer (Promega, E2810). Luminescence was measured using the FLUOstar Omega spectrophotometer and by following the manufacturer's instructions.

SARS-CoV-2 Infection Inhibition Assays
Vero-E6 cells were seeded in 24-well plates overnight before inoculation with SARS-CoV-2 at an MOI of 0.1 in the presence of 1-3 at different concentrations for 48 h at 37 • C. Vero-E6 cells were trypsinized and fixed for 20 min with 3.7% PFA. The cells were then rinsed with PBS and processed for flow cytometric assay, Cytoflex (Beckman, Villepinte, France), as previously described using monoclonal human igG1 antibody SARS-CoV-2 spike (clone H4) for 1 h, followed by a Goat anti-Human IgG Cross-Adsorbed secondary antibody, Alexa Fluor 488 (Thermo Fisher Scientific, Waltham, MA, USA), for the detection