Identification of (20R)‐protopanaxadiol from Panax ginseng as a novel anti‐SARS‐CoV‐2 compound

Abstract

COVID‐19 remains a global health concern, yet current antivirals are restricted to high‐risk patients and face issues of resistance. To establish a safer evaluation system, we constructed a noninfectious SARS‐CoV‐2 replicon using a BAC vector driven by the CMV promoter. Among the tested cell lines, HuH‐7.6c supported the most efficient replication. Replication was inhibited by remdesivir, nirmatrelvir, and molnupiravir, but not by favipiravir or AT‐527. A screening of 373 food‐derived compounds identified (20R)‐protopanaxadiol (PPD), a ginseng metabolite, as a novel inhibitor. Together, this work highlights this BAC‐vectored replicon as a practical platform for antiviral screening and identifies the potential of safe, food‐derived supplements in combating SARS‐CoV‐2.

We established a noninfectious BAC‐based SARS‐CoV‐2 replicon that enables antiviral screening under BSL‐2 conditions. Using this platform, we screened 373 food‐derived compounds and identified (20R)‐protopanaxadiol from Panax ginseng as a novel inhibitor of SARS‐CoV‐2 RNA replication, highlighting the value of safe replicon systems for antiviral discovery.

Abbreviations

BAC

bacterial artificial chromosome

BSL

biosafety level

COVID‐19

coronavirus disease 2019

EC₅₀

50% effective concentration

N protein

nucleocapsid protein

PPD

protopanaxadiol

SARS‐CoV‐2

severe acute respiratory syndrome coronavirus‐2

The coronavirus disease 2019 (COVID‐19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2), has had a serious impact on the global health and society of humanity. To reduce the risk of severe illness and mortality, several oral antiviral drugs have been developed and approved for clinical use. These include the RNA‐dependent RNA polymerase inhibitor molnupiravir and the 3CL protease inhibitor nirmatrelvir/ritonavir [1, 2, 3]. However, these drugs are primarily intended for COVID‐19 patients with mild‐to‐moderate symptoms who are not hospitalized but are at high risk of developing severe illness; they are not intended for widespread prescription to general mild cases or asymptomatic cases. Furthermore, the emergence of drug‐resistant strains due to viral mutations has been reported in association with antiviral therapies [4, 5, 6, 7]. Consequently, there is a need for the development of novel antiviral agents that are safer and can be used in a broader population. We considered that antiviral agents with high safety profiles should be developed for preventive or adjunctive use across a broad population.

Food‐derived natural products and plant extracts have long been utilized for the prevention and treatment of infectious diseases, with many reports regarding their antiviral activity. For example, fermented ginseng extract has been reported to improve survival rates and reduce viral loads in mouse models and in vitro experiments against influenza virus [8]. Furthermore, ginsenosides were also effective in suppressing infection in in vitro and in vivo models using influenza A virus (H1N1) [9]. Common dietary nutrients also exhibit antiviral activity; for example, beta‐carotene, vitamin D₂, and linoleic acid have been reported to inhibit hepatitis C virus (HCV) RNA replication [10]. More recently, Melissa officinalis (lemon balm) extract has also shown broad‐spectrum activity against influenza A, SARS‐CoV‐2, and MERS‐CoV in vitro [11]. These studies suggest that food‐derived compounds could serve as promising complementary or lead compounds for antiviral drug development. In the context of acute infectious diseases, early suppression of viral replication is critical to prevent symptom exacerbation and the spread of infection. Therefore, continuous intake of safe, food‐derived compounds with the potential to reduce viral load may be beneficial as an adjunctive or preventive strategy, particularly during the early phase of infection.

The study of antiviral compounds against SARS‐CoV‐2, however, is restricted by the requirement of biosafety level 3 (BSL‐3) facilities to handle infectious virus. In many regions, including Europe, Asia, and Australia, the mandatory use of BSL‐3 laboratories has become a major bottleneck in drug screening for many research institutions. Therefore, in this study, we established a replicon system that evaluates only the viral replication process without using infectious viruses. Research on the HCV has significantly advanced the development of anti‐HCV drugs leading to clinical applications using replicon systems [12, 13, 14, 15]. Thus, the replicon system is a powerful experimental system that combines safety and utility, enabling the analysis of viral RNA replication without producing infectious particles. In this study, we established a SARS‐CoV‐2 replicon assay system using synthetic genes constructed from database information, without utilizing the natural SARS‐CoV‐2 virus. The purpose of this study was to use this replicon to identify specific inhibitors of SARS‐CoV‐2 replication from dietary‐derived natural products.

Materials and methods

Plasmid construction

The SARS‐CoV‐2 replicon expression plasmid was constructed from 10 artificially synthesized fragments (GenScript, NJ, USA) based on the sequence of the Wuhan‐Hu‐1 strain (GenBank accession no. NC_045512). Each fragment was approximately 2.5 kb in size and contained BsaI restriction enzyme recognition sequences at both ends. After digesting the fragments with BsaI, they were grouped into sets of 3–4 fragments and cloned into the pHSG298 vector (Fragments 1–3, 4–6, 7–10). Then, the fragments were cut out from the vector again with BsaI and introduced into the bacterial artificial chromosome (BAC) vector pSMART BAC v2.0 (#42030‐2, Lucigen, WI, USA). Ligation was performed using the TaKaRa DNA Ligation Kit LONG (#6024, TaKaRa, Shiga, Japan), and transformation was carried out using BAC‐Optimized Replicator v2.0 Electrocompetent Cells (#60210‐2, Lucigen, WI, USA). A polymerase‐defective mutant (SAA; D760A/D761A substitutions in nsp12) was generated using the QuikChange Site‐Directed Mutagenesis Kit (Agilent, Santa Clara, CA, USA) according to the manufacturer's protocol [16].

Cells lines

Vero cells (derived from African green monkey kidney) were kindly provided by Prof. Eizuru, Kagoshima University. MDCK cells (derived from dog kidney) and BHK‐21 cells (derived from baby hamster kidney) were purchased from the JCRB Cell Bank (MDCK; #JCRB9029, BHK‐21; # JCRB9020). HEK293 cells (RRID:CVCL_0045; derived from human embryonic kidney), Hep3B cells (RRID:CVCL_C8Q9; derived from human hepatocellular carcinoma), and HuH‐7 cells (RRID:CVCL_0336; derived from human hepatocellular carcinoma) were kindly provided by Prof. Kato, Okayama University. HuH‐7.6c cells and RSc cells are subclones derived from HuH‐7 cells [13, 17]. All cell lines were maintained in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) (Thermo Fisher Scientific, MA, USA) at 37°C in a humidified atmosphere containing 5% CO₂. Human cell lines were obtained from established sources as described above. HuH‐7.6c and HuH‐7.RSc are laboratory‐derived cured subclones established from parental HuH‐7 cells. The parental HuH‐7 cells and both HuH‐7‐derived subclones were characterized by sequencing the full‐length coding region of the human albumin transcript, confirming their human origin. In addition, the HuH‐7‐derived subclones were confirmed to be free of mycoplasma contamination at the time of patent deposition.

Reagents and antibodies

Medicine Food Homology compound library, remdesivir, favipiravir, and MK‐4482 (EIDD‐1931) were purchased from Selleck Chemicals (TX, USA). PF‐07321332 was purchased from Amadis Chemical (Hangzhou, China), and AT‐527 was purchased from MedChemExpress (NJ, USA). The antibodies used were SARS‐CoV‐2 nucleocapsid (N) protein antibody (#33717, Cell Signaling Technology, MA, USA), β‐actin antibody (#A5441, Sigma‐Aldrich, MO, USA), and mouse IgG (#NA931, GE Healthcare, IL, USA).

Luciferase assay

The cells (2 × 10⁵ cells·well⁻¹) were seeded into a 6‐well plate. After 24 h, 2 μg of the replicon expression vector was transfected using FuGENE® HD Transfection Reagent (Promega, WI, USA). Twenty‐four hours after transfection, cells were detached with trypsin and seeded into a 24‐well plate (2 × 10⁴ cells·well⁻¹). The candidate antiviral agents were added 24 h later. The cell culture supernatants were collected 48 h after the addition of the drug. Luciferase activity was measured according to the Nano‐Glo™ Luciferase Assay (Promega) protocol.

WST‐1 cell proliferation assay

HuH‐7.6c cells (3 × 10³ cells) were seeded into 96‐well plates, and candidate antiviral agents were added after 24 h. The cytotoxicity was assessed using the WST‐1 assay according to the Premix WST‐1 Cell Proliferation Assay System protocol (Takara Bio, Shiga, Japan) 48 h after the addition of the drug.

Western blot analysis

The cells (2 × 10⁵ cells·well⁻¹) were seeded into 6‐well plate. After 24 h, 2 μg of the replicon expression vector was introduced into the cells using transfection reagent. After 24 h, the cells were detached with trypsin and seeded into 6‐well plate. Twenty‐four hours after cell seeding, the candidate antiviral drug was added. Forty‐eight hours after addition of drug, the cells were lysed by adding 2× SDS gel loading buffer, and crude protein fractions were recovered by sonication. Preparation of cell lysates, SDS/polyacrylamide gel electrophoresis (SDS/PAGE), and immunoblotting were performed according to previously described methods [18].

Quantitative RT‐PCR analysis

HuH‐7.6c cells (2 × 10⁵ cells·well⁻¹ at 6‐well plate) were cultured for 24 h before the transfection with replicon expression vector. The replicon expression vector (2 μg) was introduced into the cells using transfection reagent. After 24 h, the cells were detached with trypsin and seeded into 6‐well plate. After 24 h, the candidate antiviral drugs were added. After 48 h for treatment, total RNA was extracted using the RNeasy Mini Kit (Qiagen). Each of the RNAs were digested by RQ1 RNase‐free DNase (Promega) at 37°C for 30 min. Then, DNase was inactivated at 65°C for 10 min. Reverse transcription was performed by M‐MLV Reverse Transcriptase (Invitrogen) with Oligo(dT)12–18 primer (Invitrogen). Quantitative PCR was performed using THUNDERBIRD SYBR™ qPCR Mix (TOYOBO) on Applied Biosystems StepOnePlus Real‐Time PCR System. The ORF1a gene was detected using the primers NIID_WH‐1_F509 (5′‐CTCGAACTGCACCTCATGG‐3′) and NIID_WH‐1_R854 (5′‐CAGAAGTTGTTATCGACATAGC‐3′), which were described in a previous report [19]. The GAPDH was detected using the primers hGAPDH (5′‐ GACTCATGACCACAGTCCATGC‐3′) and hGAPDHr (5′‐GAGGAGACCACCTGGTGCTCAG‐3′). The thermal cycling conditions were as follows: 1 min at 95°C for denaturation, 35 cycles of 15 s at 95°C, 15 s at 55°C, 45 s at 72°C. Fluorescence of SYBR Green was detected in each cycle.

Statistical analysis

Data are presented as mean ± standard deviation (SD) from at least three biological replicates. P‐values were determined by Student's t‐test. *P < 0.01 was considered statistically significant.

Results

Establishment of SARS‐CoV‐2 replicons via artificial gene synthesis

To assay SARS‐CoV‐2 RNA replication safely, we constructed a SARS‐CoV‐2 replicon assay system. First, we artificially synthesized genes based on the Wuhan‐Hu‐1 strain genome sequence, in which the structural gene regions were replaced with a secreted Nano‐luciferase (sNL) gene and a neomycin resistance gene (Neo), while retaining the N gene. This design enabled quantitative measurement of RNA replication activity without producing viral particles. Furthermore, a mutant replicon with the conserved SDD motif (amino acids 759–761) of the RNA‐dependent RNA polymerase (nsp12) replaced by SAA was created as a negative control to eliminate replication‐independent background signals (Fig. 1A) [20].

Fig. 1.

Schematic of the genetic composition of the SARS‐CoV‐2 replicon. (A) The schematics of the SARS‐CoV‐2 full‐length genome (top), wild‐type replicon (WT, middle), and polymerase‐inactive mutant (SAA, bottom). The structural gene region was replaced with secreted Nano‐luciferase gene (sNL), the neomycin resistance gene (Neo), and the nucleocapsid (N). Transcription regulatory sequence in the leader (TRS‐L) and transcription regulatory sequence in the body (TRS‐B) are indicated by pink inverted triangles. SAA: The SDD motif (amino acids 759–761) of the nsp12 protein was replaced with the amino acid sequence of SAA as the negative control. (B) The schematic of the SARS‐CoV‐2 replicon expression vector. The SARS‐CoV‐2 replicon was artificially synthesized by dividing it into 10 fragments of approximately 2.5 kb. Ten synthetic gene fragments were ligated with BsaI and introduced into a bacterial artificial chromosome (BAC) vector.

To construct the vector, the SARS‐CoV‐2 replicon was artificially synthesized by dividing it into 10 fragments of approximately 2.5 kb each, with BsaI recognition sites added to both ends. These fragments were excised with BsaI and cloned by ligating 3–4 fragments together. Finally, the fragments were excised again and introduced into the BAC vector to complete the replicon expression plasmid (Fig. 1B). The resulting plasmid does not produce infectious particles, allowing for safe handling at BSL‐2 level. It can now serve as a foundation for subsequent antiviral drug screening.

Comparison of replication efficiency among different cell lines

To determine the most permissive host cells, wild‐type (WT) or polymerase inactivation variant (SAA) replicons were first introduced into multiple cell lines, and luciferase activity was measured over time. In Vero and HEK293 cells, the peak luciferase activities were observed 48 h post‐transfection, while in MDCK cells, the peak was observed at 72 h. However, in all cases, SAA showed higher activity than WT, suggesting that the signal was not replication dependent. In Hep3B cells, both WT and SAA showed a time‐dependent increase in activities, but no differences were observed between them. Therefore, SARS‐CoV‐2 RNA replication was not detected in these cells. In contrast, BHK‐21 and HuH‐7 cells exhibited a clear difference in luciferase activity between WT and SAA. Notably, HuH‐7 cells showed time‐dependent and high‐level luciferase activity, suggesting suitability for SARS‐CoV‐2 RNA replication (Fig. 2A). Furthermore, replication efficiency was compared using subclones derived from HuH‐7 cells.

Fig. 2.

Comparison of SARS‐CoV‐2 replicon replication activity in various cell lines. (A) Wild‐type (WT) or polymerase‐inactivating mutant (SAA) SARS‐CoV‐2 replicons were transfected into Vero, HEK293, MDCK, Hep3B, BHK‐21, and HuH‐7 cells. Luciferase activity was measured 24–96 h post‐transfection. Red circles indicate WT; gray squares indicate SAA. (B) WT or SAA replicons were transfected into HuH‐7, HuH‐7.RSc, and HuH‐7.6c cells. Luciferase activity was measured 24–96 h post‐transfection. (C) After transfection of WT or SAA replicons into HuH‐7.6c cells, luciferase activity was evaluated (top) and nucleocapsid (N) protein expression was detected (bottom) at 24, 48, 72, and 96 h. β‐Actin was used as a control for the amount of protein loaded into each lane. Data are presented as mean ± SD from three biological replicates. Error bars indicate SD. Statistical significance was determined using Student's t‐test. *P < 0.01.

The subclone cells used were HuH‐7.RSc and HuH‐7.6c. These cured cells were generated by eliminating HCV from cells that originally harbored the virus. When WT or SAA replicons were introduced into HuH‐7.RSc and HuH‐7.6c, significant differences were observed between WT and SAA in all subclones, demonstrating higher replication efficiency than the parental strain. Notably, HuH‐7.6c exhibited a marked increase in luciferase activity over time, confirming it as the cell line most efficiently maintaining replication (Fig. 2B). Moreover, evaluation of replicon‐derived protein expression in HuH‐7.6c cells revealed that N protein expression was detected 48 h or later after WT replicon transfection, whereas no expression was detected with the SAA replicon (Fig. 2C). These results suggested that HuH‐7.6c cells are the optimal host cells for SARS‐CoV‐2 replicon analysis, and they were used in subsequent experiments.

Validation of the replicon using clinically approved antivirals

To evaluate whether this system can reproduce the activity of antiviral drugs, we introduced replicons into HuH‐7.6c cells and validated the utility of the replicon system using approved drugs. Remdesivir, the RNA‐dependent RNA polymerase inhibitor, reduced luciferase activity in a concentration‐dependent and time‐dependent manner between 24 and 96 h, with an EC₅₀ of 16.4 nm (Fig. 3A). This confirmed that the replicon system can sensitively reflect the antiviral effect of the drug.

Fig. 3.

Evaluation of replicon using approved drugs and natural product libraries. (A) HuH‐7.6c cells were transfected with replicon, and remdesivir was added 24 h later. The luciferase activity was evaluated from 24 to 96 h after addition when remdesivir was treated at concentrations of 10, 20, and 100 nm, respectively (left). The luciferase activity was evaluated at 48 h after addition of serially diluted remdesivir, and the 50% effective concentration (EC₅₀) value was calculated (right). (B) HuH‐7.6c cells were transfected with the replicon, and approved antiviral drugs were added 24 h later. The nucleocapsid (N) protein expression (top) and the luciferase activity (bottom) were evaluated 48 h after addition of drug. β‐Actin was used as a control for the amount of protein loaded per lane. (C) HuH‐7.6c cells were transfected with replicon. The compound 373 from the food‐derived natural product library was added 24 h post‐transfection. It was diluted to a final concentration of 10 μm. The luciferase activity was measured 48 h later. The untreated group was indicated as (−), and remdesivir (100 nm) was used as a positive control. (D) HuH‐7.6c cells were treated with compound 373. It was diluted to a final concentration of 10 μm. The cell viability was assessed using the WST‐1 cell proliferation assay after 48 h. Data are presented as mean ± SD from three biological replicates. Error bars indicate SD. Statistical significance was determined using Student's t‐test. *P < 0.01.

Furthermore, the 3CL‐protease inhibitor PF‐07321332 (nirmatrelvir) and the polymerase inhibitor MK‐4482 (molnupiravir) also markedly suppressed replication. In contrast, favipiravir and AT‐527 had no inhibitory effect. The expression of N protein and ORF1a gene was also consistent with the luciferase assay results, decreasing with remdesivir, PF‐07321332, and MK‐4482, but not with favipiravir and AT‐527 (Fig. 3B; Fig. S1). Drugs with confirmed clinical efficacy were shown to suppress viral RNA replication in the replicon system, whereas favipiravir and AT‐527, whose development was discontinued owing to insufficient clinical efficacy, showed no inhibitory activity. Thus, the replicon system accurately predicted the clinical efficacy profiles of clinically used antivirals.

Screening of food‐derived natural‐product compounds

Next, we screened for antiviral candidates using a library of food‐derived natural products. After transfection of HuH‐7.6c cells with the replicon, we added 373 compounds and evaluated luciferase activity after 48 h. Luciferase assays identified 10 compounds that suppressed replication by more than 50% compared to the untreated group (Fig. 3C; Table 1). In parallel cell viability assays, six of these compounds were found to maintain survival rates of 80% or higher (Fig. 3D). Among the six compounds, Daurisoline inhibited luciferase activity by 85.9% with the cell viability of 99.7%. Next, (20R)‐protopanaxadiol (PPD) inhibited luciferase activity by 70% with the cell viability of 85.6%. Daurisoline has previously been reported to exhibit antiviral activity against SARS‐CoV‐2 [21]. Therefore, (20R)‐PPD was selected as a promising antiviral candidate compound against SARS‐CoV‐2 (Table 1).

Table 1.

Identification of antiviral candidate compounds against SARS‐CoV‐2 replication.

	Compound name	Relative secNLuc activity (%)	Cell viability (%)
	(−)	100	100
	Remdesivir	22.8	134.1
1.	Shikonin	0.2	40.0
2.	Periplocin	1.6	34.4
3.	Daurisoline	14.1	99.7
4.	Piperlongumine	29.6	19.3
5.	(20R)‐Protopanaxdiol	30.0	85.6
6.	Protodioscin	31.9	28.1
7.	Dehydrodiisoeugenol	38.9	106.7
8.	Dauricine	40.7	114.3
9.	(S)‐2‐Hydroxysuccinic acid	42.2	119.7
10.	Panaxatriol	49.3	158.6

Characterization of (20R)‐protopanaxadiol

We examined in detail the antiviral effects of (20R)‐PPD by screening and identifying compounds. The optical isomers, the (20R)‐ and (20S)‐PPD, are shown in Fig. 4A. We investigated whether the structural difference affects antiviral activity.

Fig. 4.

Evaluation of the antiviral activity of (20R)‐protopanaxadiol [(20R)‐PPD]. (A) Chemical structures of (20R)‐protopanaxadiol [(20R)‐PPD] and (20S)‐protopanaxadiol [(20S)‐PPD]. (B) HuH‐7.6c cells were transfected with replicons, and (20R)‐PPD or (20S)‐PPD was added 24 h later. The luciferase activity was measured after 48 h. The 50% effective concentration (EC₅₀) values were calculated for (20R)‐PPD (left). HuH‐7.6c cells were treated with (20R)‐PPD. The cell viability was assessed using the WST‐1 cell proliferation assay after 48 h. The 50% cytotoxic concentration (CC₅₀) values were calculated for (20R)‐PPD (right). (C) HuH‐7.6c cells were transfected with the replicon, and (20R)‐PPD or (20S)‐PPD were added 24 h later. The nucleocapsid (N) protein expression (top) and luciferase activity (bottom) were evaluated 48 h after addition of drug. β‐Actin was used as a control for the amount of protein loaded per lane. Data are presented as mean ± SD from three biological replicates. Error bars indicate SD. Statistical significance was determined using Student's t‐test. *P < 0.01.

The (20R)‐PPD inhibited replication in a dose‐dependent manner, exhibiting an EC₅₀ value of 5.46 μm and a CC₅₀ value of 26.7 μm (Fig. 4B). Furthermore, (20R)‐PPD significantly repressed the expression of N protein and the ORF1a gene. These results were consistent with the luciferase assay (Fig. 4C; Fig. S2). In contrast, the antiviral activity of (20S)‐PPD was very weak compared to (20R)‐PPD (Fig 4B,C; Fig. S2). Collectively, our findings demonstrate that the BAC‐based SARS‐CoV‐2 replicon represents a robust platform suitable for operation under BSL‐2 conditions. It not only reproduces the antiviral profiles of approved drugs but also enables the efficient identification of novel inhibitors such as the ginseng‐derived metabolite (20R)‐PPD.

Discussion

In this study, we established a SARS‐CoV‐2 replicon system using the BAC vector that enables quantitative assessment of viral RNA replication under BSL‐2 conditions. A comparison of multiple cell lines revealed that HuH‐7.6c cells exhibited the highest replication efficiency, confirming them as a suitable host cell for antiviral drug evaluation. This replicon system faithfully reproduced the antiviral activity of clinically approved drugs and identified (20R)‐PPD as a novel antiviral candidate compound against SARS‐CoV‐2.

The replicon developed in this study offers several advantages over previously reported systems. Most existing SARS‐CoV‐2 replicons utilize the T7 promoter, which requires in vitro transcription followed by the electroporation of RNA into target cells [22, 23, 24, 25]. While these systems are effective for short‐term assays, their luciferase activity peaks 24–48 h after electroporation before declines, thereby limiting the evaluation of compounds that act in the later stages of replication. In contrast, our BAC‐based replicon driven by the CMV promoter can be introduced into cells via simple transfection, eliminating the need for in vitro RNA synthesis and electroporation. Furthermore, replication is sustained for at least 4 days, providing a longer time window for compound testing. Several groups have reported replicons using BAC vectors. While these systems improved stability and ease of handling, they lacked polymerase‐deficient controls or verification of protein expression from the replicon [26, 27]. In this study, the use of SAA mutants as replication negative controls enabled clear differentiation between background signals and true replication signals. Additionally, western blot analysis of the N protein confirmed the functionality of the replicon at the protein level. These features enhance the system's reliability, making it suitable for detailed quantitative analysis. In addition to replicon systems, minigenome‐based evaluation systems have also been reported recently [28].

We evaluated multiple cell lines and identified Huh‐7.6c cells as the most efficient host for SARS‐CoV‐2 replicon replication. The observation that subclones derived from HuH‐7 cells exhibit higher replication efficiency than their parental strain has also been reported in HCV replicon studies [29, 30]. Genetic or epigenetic alterations within these subclones likely contribute to a state permissive for RNA virus replication. Our findings suggest that similar host factors may also influence the activity of SARS‐CoV‐2 replicons as well.

Validation of this system using approved drugs confirmed its ability to accurately reflect clinically significant antiviral effects. The RNA‐dependent RNA polymerase inhibitor remdesivir, the 3CL protease inhibitor PF‐07321332 (nirmatrelvir), and the nucleoside analog MK‐4482 (molnupiravir) all exhibited potent, dose‐dependent inhibition of viral replication. In contrast, favipiravir and AT‐527 showed little to no activity, which is consistent with clinical and virological data reporting limited efficacy for these compounds [31, 32, 33]. These results confirm the robustness of the replicon system in predicting pharmacological effects observed in patients and demonstrate its applicability for drug screening. This study has several limitations. The replicon system evaluates viral RNA replication but does not assess viral entry, assembly, or release of infectious particles. Therefore, antiviral candidates identified using this platform should be further validated in infectious virus and in vivo models. Nevertheless, the present system reproduced the clinical efficacy profiles of approved antivirals, including the lack of activity of favipiravir and AT‐527, suggesting that host cell selection may substantially influence the clinical relevance of antiviral evaluation systems.

We screened food‐derived compounds using our replicon system and identified (20R)‐PPD, which is a metabolite of Panax ginseng ginsenosides, as a novel inhibitor of SARS‐CoV‐2 replication. Korean ginseng (Panax ginseng) is widely used in traditional East Asian medicine and modern supplements due to its immunomodulatory, antioxidant, and anti‐inflammatory effects [34, 35]. Ginsenosides and their metabolites are known to exhibit antitumor, anti‐inflammatory, and immunomodulatory effects. Reported mechanisms of action include the induction of caspase‐dependent apoptosis, the reduction of inflammatory cytokine production via suppression of the NF‐κB pathway, and the inhibition of reactive oxygen species production [36, 37]. Previous reviews also support the efficacy of Korean ginseng in respiratory and other infections, suggesting its usefulness as an adjunctive therapy through immunomodulatory and anti‐inflammatory effects [38]. Furthermore, Korean Red Ginseng extract has been shown to reduce inflammation and oxidative stress in a SARS‐CoV‐2 pseudovirus model, demonstrating its potential as an effective antiviral functional food [39]. The results of this study support these findings and further demonstrate that the ginseng metabolite, (20R)‐PPD, inhibits SARS‐CoV‐2 replication. In addition to the discovery of (20R)‐PPD, our findings suggest that food‐derived compounds exhibit potential as a safe and readily available source of antiviral agents. While their efficacy may be lower than that of direct‐acting antivirals, continuous dietary intake or supplementation may contribute to reducing viral load and inflammation during the early stages of infection or following exposure. Therefore, they could be incorporated into preventive or adjunctive treatment strategies. In conclusion, as the history of replicon research demonstrates, noninfectious replication systems have significantly advanced antiviral drug development. The SARS‐CoV‐2 replicon system established in this study is also expected to serve as a powerful platform for screening novel antiviral compounds. As the SARS‐CoV‐2 pandemic continues globally, and in light of the limitations of existing drugs and the growing risk of resistance, the development of antiviral agents derived from safe, daily ingestible supplements, and food components is critically important for future infectious disease control.

In this study, we identified natural compounds that exhibit potent antiviral activity against SARS‐CoV‐2. These findings highlight the potential of safe, food‐derived substances as adjunctive options for COVID‐19 therapy. Further investigations, including in vivo and clinical studies, are warranted to validate their efficacy and to support the development of affordable antiviral nutraceuticals for future pandemic preparedness.

Conflict of interest

The authors declare that they have no conflict of interest.

Author contributions

MT and MI conceived and designed the study. MT performed the experiments, analyzed and interpreted the data, and prepared the figures. NM and RU contributed to experimental investigation and data acquisition. MT drafted the manuscript. NT and MI reviewed and edited the manuscript. MI supervised the study, administered the project, and acquired funding.

Supporting information

Fig. S1. Evaluation of replicon using approved drugs. HuH‐7.6c cells were transfected with the replicon, and approved drugs were added 24 h later. The quantitative analysis of ORF1a by quantitative reverse transcription PCR (qRT‐PCR) (top) and luciferase activity (bottom) were evaluated 48 h after addition of drug. Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) was used as an internal control. Data are presented as mean ± SD from three biological replicates. Error bars indicate SD.

Fig. S2. Evaluation of the antiviral activity of (20R)‐protopanaxadiol [(20R)‐PPD]. HuH‐7.6c cells were transfected with replicons, and (20R)‐PPD or (20S)‐protopanaxadiol [(20S)‐PPD] was added 24 h later. The quantitative analysis of ORF1a by quantitative reverse transcription PCR (qRT‐PCR) (top) and luciferase activity (bottom) were evaluated 48 h after addition of drug. Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) was used as an internal control. Data are presented as mean ± SD from three biological replicates. Error bars indicate SD.

Data accessibility

The data analyzed in this study are included in the paper. Data will be made available upon request.