MDCK (Madin-Darby canine kidney) cells (ATCC, Manassas, VA, USA, CCL-34), and HEK293T (Human embryonic kidney) cells (ATCC, CRL-3216) were cultured at 37°C, 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% heat-inactivated fetal bovine serum (Gibco) and 1% PSG (100 U/mL penicillin, 100 μg/mL streptomycin). A549 (Human lung epithelial carcinoma) cells (ATCC, CCL-185) were cultured at 37°C, 5% CO₂ in F-12K Nutrient Mixture (F-12K, Thermo Fisher Scientific) supplemented with 10% heat-inactivated fetal bovine serum (Gibco) and 1% PSG (100 U/mL penicillin, 100 μg/mL streptomycin). H1N1 subtype IAV A/California/07/2009 (CA07) were propagated in MDCK cells. The viruses carrying the gene of green fluorescent protein (GFP) used in this experiment were provided by the Key Laboratory of Pathogens and Biosecurity at the Academy of Military Medical Sciences. All the Chinese herbal compounds needed for the experiments were purchased from MCE.

Well-grown MDCK cells were inoculated at 250 μL of medium per well into a 48-well plate. After ensuring uniform cell distribution, the plate was gently transferred to a 37°C, 5% CO₂ incubator for continued cultivation. When the cell confluence reached over 80%, the medium in each well was aspirated and replaced with virus isolation serum-free medium. A 200 μL mixture of GFP-labeled influenza virus (Sun et al., 2014) was added and incubated for 1 h at an multiplicity of infection (MOI) of 0.001. The plate was shaken every 15 minutes to prevent cell clumping. Compounds were diluted to 50 μM and prepared as 250 μL solutions in the plate, with one duplicate well for each compound. The plates were then cultured again in the 37°C, 5% CO₂ incubator. At 12, 18, and 24 h post-infection, cell status and GFP fluorescence intensity were observed using Gen5 fluorescence microscopy. The viral infection-induced GFP expression and its changes under drug treatment were analyzed. Fluorescence intensity measurements were employed to evaluate viral infectivity and drug efficacy against the infection. Preliminary screening identified compounds with significant antiviral activity.

Total RNA was extracted from virus culture supernatant using a PureLink RNA Mini Kit (Thermo Fisher Scientific). qRT-PCR was performed using a One Step TB Green PrimeScript PLUS RT-PCR Kit (TaKaRa, Shiga Prefecture, Kusatsu City, Japan) with a Light Cycler 480 II (Roche, Basel, Switzerland). Virus titer was determined by qRT-PCR and converted to plaque-forming units (PFU) numbers by standard curve.

MDCK cells were cultured at 500 μL of medium per well in a 24-well plate. After achieving uniform cell coverage, the plates were transferred to a 37°C, 5% CO₂ incubator for continued culture. When cell confluence exceeded 80%, the original medium was aspirated and replaced with serum-free viral isolation medium. 200 μL of wild-type CA07 virus was added and incubated at an MOI of 0.001 for 1 h. The plates were shaken every 15 minutes to prevent clumping. The compounds were diluted to 50 μM and prepared as 1800 μL of solutions, with three replicate wells per compound. The plates were then returned to the incubator for further culture. At 12, 24, 48, and 72 h post-infection, cell viability was assessed under a microscope, and 230 μL of supernatant was collected for nucleic acid extraction and qRT-PCR to evaluate antiviral efficacy. Viral replication kinetics experiments confirmed that the four compounds exhibited antiviral activity. A549 cells were cultured using F-12K medium at an MOI of 0.01, following the same protocol.

MDCK and A549 cells were seeded in 96-well plates and incubated at 37°C with 5% CO₂ for over 12 h. After confirming cell adhesion, the plates were washed three times with PBS, followed by the addition of 100 μL of virus-isolated serum-free medium containing the compound at 200 μM starting concentration. Six duplicate wells were prepared for each concentration, with one column containing only medium as the blank control and another containing DMSO as the negative control. The plates were incubated at 37°C for 24 h, then allowed to equilibrate at room temperature for 30 minutes. Two hours prior, the CellTiter-Glo Reagent (Promega) was thawed, and 100 μL of substrate-buffer mixture was added per well. After mixing, the plates were incubated on a shaker for 2 minutes to lyse cells. The plates were then incubated at room temperature for 10 minutes to stabilize luminescence signals, followed by automated detection. Cell survival rate (%) = (the absorbance of drug-treated group/the absorbance of blank control group) × 100%.

293T cells were prepared for 24 h in advance. Plate them in 48-well plates and incubate overnight at 37°C with 5% CO₂. When cells reach 80% confluence, they are then transfected with Lipofectamine 3000 transfection kit to introduce IAV PB1, PB2, PA, NP, pHH21-Fluc, and pRL-TK plasmids. After a 5-minute incubation separately, liposomes and genetic material were mixed and incubated for more than 15 minutes, were supplemented with 25 μL of DNA-liposome complex per well, and then were incubated at 37°C with 5% CO₂ for 6 h. Finally, the cell supernatant was removed and the cells were updated with compound-containing medium. Six gradient concentrations (50 μmol/mL total) were prepared using three sub-wells per gradient, with Baloxavir marboxil as positive control, and with DMSO and blank as negative controls. Cells for all groups were maintained at 37°C with 5% CO₂ for 24 h, and then both Firefly luciferase (Fluc) and Renilla reniformis (Rluc) were detected using Promega Dual-Luciferase Reporter Assay System kit, and the Fluc/Rluc was calculated.

The A549 cells were inoculated into a 6-well plate at a density of 2 mL of medium per well. After the cells were evenly spread, they were transferred to a 37°C, 5% CO₂ incubator for further culture. When the cell density reached 80-90%, the cells were washed three times with PBS and replaced with serum-free medium. A 500 μL suspension of wild-type CA07 virus was added, and the cells were incubated at an infection multiplicity (MOI = 0.1) for 1 h. The cells were shaken every 15 minutes to prevent clumping. The compound was diluted to the corresponding concentration and prepared into a 7 mL solution, with concentrations of Andrographolide (50 μM), Cryptotanshinone (6.25 μM), Aloe emodin (25 μM), Emodin (25 μM), and DMSO as the negative control. Three replicate wells were set for each compound, and the culture plates were returned to the incubator for continued cultivation. After an incubation for 24 h, the cell status was observed, and the medium was aspirated. TRIzol^® reagent (Thermo Fisher Scientific) was used to extract total RNA for transcriptome analysis. The medium was removed, and the cells were washed with PBS before being digested with trypsin to collect proteins for proteome sequencing analysis. For all the above compounds, a control group without wild-type CA07 virus was established, using medium containing 10% FBS, with other procedures remaining unchanged. The transcriptome and proteome sequencing analyses were performed using GENE DENOVO software (China Guangzhou). Functional annotation and enrichment analysis were performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database’s biological processes (BP), cellular components (CC), and molecular functions (MF) classifications.

The GraphPad Prism 9.0 software was used to conduct all statistical analysis. Data were presented as means ± SD of at least three independent experiments. Differences between the two groups were evaluated using Student’s t-test, and differences between three or more groups were evaluated using analysis of variance (ANOVA). statistical significance was set at p < 0.05.

Natural compound libraries often contain hundreds of natural compounds with diverse functions. To improve screening efficiency and narrow the screening scope, we adopted a method combining literature mining, compound database screening, and virtual screening based on structural similarity, supplemented by prior knowledge for manual precision screening. First, we selected the representative traditional medical literature, the Pharmacopoeia of the People’s Republic of China, as the target text. Through literature mining, compound database screening, virtual screening based on structural similarity, and comparative analysis with known natural chemical compounds exhibiting antiviral activity against influenza viruses, we identified 50 representative compounds with potential antiviral activity from hundreds of TCM materials listed in the Pharmacopoeia. Subsequently, we integrated prior knowledge for manual precision screening and referenced the antiviral activity annotations of compounds on the MCE website to clarify the screening direction. The primary screening targets were compounds with activity related to “antiviral signaling pathways,” including those capable of modulating the interferon (IFN) pathway, NF-κB signaling pathway, JAK-STAT signaling pathway, MAPK signaling pathway, or natural chemical compounds reported to exhibit antiviral activity against other RNA viruses (e.g., RSV, HCV, SARS-CoV-2, EV71, etc.) (Figure 1A). After this rigorous stepwise screening process, we ultimately identified 20 compounds with potential IAV activity as candidate compounds (Supplementary Table 1). To evaluate the antiviral activity of selected compounds against IAV, MDCK cells were infected with GFP-IAV and added 50 μM of the natural compound to the supernatant, using DMSO as the negative control and oseltamivir as the positive control. In this screening system, GFP fluorescence intensity directly reflects viral replication within cells, and stronger GFP fluorescence intensity indicates higher viral replication efficiency (Nie et al., 2024). The detection results 24 h after viral infection showed that compared with the DMSO negative control group, the GFP fluorescence intensity in the Aloe emodin, Cryptotanshinone, Emodin, and Andrographolide-treated groups was significantly reduced (Figure 1B), indicating that these four compounds could effectively inhibit the replication of IAV in MDCK cells, preliminarily confirming their anti-IAV activity. The chemical structures of the four aforementioned active compounds are illustrated in Figure 1A. In summary, through a multi-tiered screening system of “preliminary virtual screening-manual precision screening-GFP reporter virus validation”, we have preliminarily identified four natural compounds with confirmed anti-IAV activity: Aloe emodin, Cryptotanshinone, emodin, and Andrographolide. These compounds provide important candidate molecules for further investigation of their antiviral mechanisms and the development of novel anti-IAV drugs.

To further validate antiviral efficacy of these four compounds against IAV, MDCK cells were subsequently infected with wild-type CA07 virus and added natural compounds Aloe emodin, Cryptotanshinone, emodin, and Andrographolide to the supernatant. Oseltamivir was used as the positive control and DMSO as the negative control. Virus titers were measured at different time points (12, 24, 48, or 72 h post-infection) using qRT-PCR. Results showed that the viral replication efficiency in cells treated with the four compounds was significantly lower than that in the DMSO group (Figure 2A). To further confirm their antiviral activity in different cell lines and determine optimal concentrations, experiments were conducted using MDCK cells (MOI = 0.001) and A549 cells (MOI = 0.01) infected with wild-type CA07 virus. After 1-hour adsorption, the four compounds were diluted in a 2-fold concentration gradient (200, 100, 50, 25, 12.5, 6.25 and 3.125 μM) and then incubated with infected cells. Oseltamivir served as the positive control, while DMSO as the negative control. qRT-PCR was employed to measure viral titters at each of the four time points (12, 24, 48, or 72 h post-infection). qRT-PCR results showed that the compounds significantly reduced IAV replication in both MDCK and A549 cells compared with the DMSO control (Figures 2B, C). At higher concentrations (100 and 200 µM), however, potential effects on cell viability were observed and require further validation. The above results indicated that these four compounds have an inhibitory effect on IAV proliferation, but the specific mechanisms involved require further investigation.

To evaluate the cytotoxicity of these compounds, researchers treated MDCK and A549 cells with a starting concentration of 100 μM, which was then diluted twofold. Each experimental group included a blank control and a DMSO-negative control. Cell viability was assessed after 24 h (Figures 3A, B), and the median cell lethal concentration (CC50) for the four compounds was calculated (Figures 3C, D). The results showed that in MDCK cells, the CC50 values for Aloe emodin, Cryptotanshinone, Emodin, and Andrographolide were 11.48 μM, 7.525 μM, 13.59 μM, and 16.74 μM, respectively (Figure 3C). In A549 cells, the CC50 values were 23.92 μM, 6.503 μM, 31.11 μM, and 64.94 μM, respectively (Figure 3D). These compounds exhibited cell-type-dependent toxicity. Further analysis revealed that Cryptotanshinone demonstrated the strongest cytotoxicity in both cell types, with CC50 values below 8 μM. Notably, its toxicity in A549 cells (6.503 μM) was slightly higher than in MDCK cells (7.525 μM), making it the only component among the four compounds that showed superior toxicity in tumor cells (A549) compared to normal epithelial cells (MDCK). In contrast, Andrographolide exhibited the weakest cytotoxicity, with a CC50 value as high as 64.94 μM in A549 cells, which was 3.88 times higher than its CC50 value in MDCK cells, demonstrating the most significant cell type-dependent difference. Aloe emodin and Emodin shared similar toxicological profiles, both showing stronger toxicity to normal MDCK cells.

To investigate whether the antiviral effects of compounds were related to inhibition of IAV polymerase activity, the viral polymerase activity was detected by a dual-luciferase reporter assay system with compounds treated. As baloxavir is known to inhibit IAV polymerase activity, it was selected as a positive control. The pHW2000 plasmids encoding PB2, PB1, PA, and NP proteins, and the pHH21 plasmids expressing negative vRNA-like firefly luciferase (Fluc) RNA were co-transfected into 293T cells, with pRL-TK plasmids expressing renilla luciferase as an internal reference. Luciferase activity was measured 24 h after transfection. Combined with viral replication kinetics experiments and cellular activity assays, the results showed that Emodin, and Andrographolide had no significant effect on IAV polymerase activity (Figure 3E). At low concentrations (below 6.25 μM), neither Aloe emodin nor Cryptotanshinone had any effect on polymerase activity. However, at high concentrations (above 12.5 μM), Aloe emodin exhibited inhibitory capacity, while Cryptotanshinone, conversely, demonstrated promoting capacity (Figure 3E). To further confirm whether these two compounds directly affect luciferase expression, rather than through the IAV polymerase, additional validation experiments were conducted. 293T cells were transfected with pcDNA3.1-Fluc and pRL-TK and simultaneously treated with different concentrations of Aloe emodin or Cryptotanshinone. The experimental results demonstrated that both compound groups exhibited elevated fluc expression levels at high concentrations (Figure 3F). These findings collectively suggest that Aloe emodin may exert inhibitory effects on the IAV polymerase itself. However, the mechanism of action of Cryptotanshinone appears to be independent of IAV polymerase and may instead function through nonspecific regulatory mechanisms, further revealing its potential targets within the host transcription-translational regulatory network—for instance, by enhancing host Pol II-mediated transcriptional activity, modulating the efficiency of host translation mechanisms, and regulating cellular states and stress responses. The absence of such IAV polymerase inhibitory effects in the other two compounds indicates that their antiviral effects may originate from alternative host signaling pathways. These findings provide critical insights for elucidating the underlying mechanisms and guiding subsequent application development.

To elucidate the antiviral mechanisms of these four compounds against IAV, we conducted transcriptomic and proteomic sequencing analyses to detect cellular changes after compound treatment. A549 cells were either treated with DMSO as a mock or exposed to the four compounds for 24 h, followed by extraction of total RNA and total protein for subsequent transcriptomic and proteomic sequencing. The omics analysis results demonstrated that all four compounds significantly influenced the expression of virus infection-related genes in A549 cells. Compared to the control group, volcano plots (Figures 4A–D, E–H) revealed that each compound-treated group exhibited a significant number of differentially expressed genes and proteins. Notably, the most pronounced transcriptional changes were induced by Cryptotanshinone (Figure 4B) and Emodin (Figure 4C), with the number and distribution range of red (significantly upregulated) and blue (significantly downregulated) dots in the volcano plots exceeding those of Aloe emodin and Andrographolide. Transcriptomic analysis revealed that all compounds collectively enriched pathways including cancer, viral infections, and immune system (Supplementary Figure 1). The effects of Aloe emodin were associated with the PI3K-Akt signaling pathway and natural killer cell-mediated cytotoxicity (Figure 5A; Supplementary Figure 1A). Cryptotanshinone involved a broader network, including TNF, MAPK/PI3K-Akt signaling pathway, and complement cascade (Figure 5B; Supplementary Figure 1B). The effects of Emodin were highly concentrated in the TNF, NF-κB, and MAPK inflammatory pathways (Figure 5C; Supplementary Figure 1C). Andrographolide was found to exert multi-target regulation via the TNF, NF-κB, MAPK, and PI3K-Akt signaling pathways (Figure 5D; Supplementary Figure 1D). Proteomic results validated and expanded these findings. Aloe emodin and Emodin were associated with IAV-related processes and the PI3K-Akt/MAPK signaling pathway (Figures 5E, G; Supplementary Figures 1E, G). Cryptotanshinone exhibited close association with the NF-κB/MAPK pathway and viral protein-cytokine interactions (Figure 5F; Supplementary Figure 1F). Andrographolide demonstrated a unique regulatory pattern, likely mediated through C-type lectin receptors and the JAK-STAT signaling pathway (Figure 5H; Supplementary Figure 1H). In summary, these four compounds can reshape cellular antiviral states by modulating the expression of innate immune and inflammatory response-related genes and proteins in host cells, thereby laying the foundation for subsequent resistance to IAV infection. Among them, activating host innate immunity represents their common antiviral strategy, while each compound exhibits distinct regulatory preferences.

Comprehensive multi-omics analysis of IAV-infected A549 cells treated with four compounds revealed specific roles of these compounds in reprogramming the host’s response to viral attack. Unlike the basal regulatory effects observed in uninfected cells, this analysis captured dynamic, infection-specific remodeling of cellular pathways by each compound. Transcriptomic analysis of infected cells demonstrated that Aloe emodin, Cryptotanshinone, Emodin, and Andrographolide induced significant and widespread changes in host gene expression post-infection (Figures 4I–L). Pathway enrichment analysis indicated that all compounds significantly regulated core antiviral defense gene networks, including pathways annotated as viral infectious diseases, immune system, and signal transduction (Supplementary Figure 2). Notably, each compound exhibited unique regulatory characteristics: Aloe emodin primarily affected IAV, TNF/NF-κB, and PI3K-Akt/MAPK pathways (Figure 5I; Supplementary Figure 2A); Cryptotanshinone was significantly enriched in IAV, TNF, and MAPK signaling cascades (Figure 5J; Supplementary Figure 2B); Emodin was associated with TNF, NF-κB, MAPK, and Toll-like receptor pathways (Figure 5K; Supplementary Figure 2C); and Andrographolide influenced IAV pathways, viral protein-cytokine receptor interactions, and the integrated TNF/NF-κB/MAPK axis (Figure 5L; Supplementary Figure 2D). Proteomic analysis of identical infection-treatment scenarios confirmed and expanded these findings, revealing significant changes in protein abundance across key functional categories (Figures 4M–P). The data further highlighted compound-specific regulation of critical infection response pathways. Aloe emodin and Emodin were associated with the modulation of NOD-like receptor signaling and the JAK-STAT/NF-κB pathway (Figures 5M, O, Supplementary Figures 2E, G). Cryptotanshinone was specifically linked to viral protein-cytokine receptor interactions, in addition to its involvement in NOD-like receptor signaling (Figure 5N; Supplementary Figure 2F). Andrographolide demonstrated its participation in both NOD-like receptor signaling and the TNF/NF-κB/MAPK network (Figure 5P; Supplementary Figure 2H). Collectively, multi-omics analysis of infected cells indicated that these four anti-IAV compounds not only regulate static host pathways but also actively and differentially reconstruct the host’s transcriptional and translational landscapes in response to IAV infection. Each compound targets distinct subpopulations of core antiviral and immunomodulatory pathways, highlighting their unique host-directed mechanisms in pathophysiological contexts.