Honeysuckle (Lonicera japonica) and Huangqi (Astragalus membranaceus) Suppress SARS-CoV-2 Entry and COVID-19 Related Cytokine Storm in Vitro

Abstract

COVID-19 is threatening human health worldwide but no effective treatment currently exists for this disease. Current therapeutic strategies focus on the inhibition of viral replication or using anti-inflammatory/immunomodulatory compounds to improve host immunity, but not both. Traditional Chinese medicine (TCM) compounds could be promising candidates due to their safety and minimal toxicity. In this study, we have developed a novel in silico bioinformatics workflow that integrates multiple databases to predict the use of honeysuckle (Lonicera japonica) and Huangqi (Astragalus membranaceus) as potential anti-SARS-CoV-2 agents. Using extracts from honeysuckle and Huangqi, these two herbs upregulated a group of microRNAs including let-7a, miR-148b, and miR-146a, which are critical to reduce the pathogenesis of SARS-CoV-2. Moreover, these herbs suppressed pro-inflammatory cytokines including IL-6 or TNF-α, which were both identified in the cytokine storm of acute respiratory distress syndrome, a major cause of COVID-19 death. Furthermore, both herbs partially inhibited the fusion of SARS-CoV-2 spike protein-transfected BHK-21 cells with the human lung cancer cell line Calu-3 that was expressing ACE2 receptors. These herbs inhibited SARS-CoV-2 M^pro activity, thereby alleviating viral entry as well as replication. In conclusion, our findings demonstrate that honeysuckle and Huangqi have the potential to be used as an inhibitor of SARS-CoV-2 virus entry that warrants further in vivo analysis and functional assessment of miRNAs to confirm their clinical importance. This fast-screening platform can also be applied to other drug discovery studies for other infectious diseases.

Keywords: COVID-19; Huangqi; SARS-CoV-2; honeysuckle; let-7a; miR-148b; microRNA; mir-146a.

FIGURE 1

The schematic for predicting TCM candidates and the workflow of the analysis of their mechanisms of action. Identification of potential TCM candidates for COVID-19 treatment. The target genes indicate the biological function affected by the drug treatment and viral mechanisms. Therefore, small-molecule reagents, viral mechanisms, and the TCM candidates can be linked through the common shared target genes. The flow chart describes the steps in the bioinformatics analysis. (A) The 38 well-known small-molecule reagents for COVID-19 treatments were obtained from UpToDate and DynaMed, and 25 out of 38 drugs had the corresponding drug targets via CLUE (TOUCHSTONE). (B) A total of 30 target genes could be obtained from 25 drugs. (C) Forty-nine target genes were identified by reviewing the COVID-19-related ARDS mechanisms and were included along with the abovementioned 30 target genes for further analysis. Two target genes, TNF and TLR7, were present in both target gene groups. (D) The target above genes were queried separately using SymMap to retrieve their connected TCMs. The ten most frequent target genes were connected to more than 200 TCMs. (E) As a result, 66 TCM candidates shared common connections with all of these 10 target genes. (F) The main goal of our big data analysis was to link the small-molecule agents and TCM via biological target genes.

FIGURE 2

The potential mechanism of Honeysuckle (Lonicera japonica) and Huangqi (Astragalus membranaceus) via the bioinformatics workflow. (A) According to the results of the FDR-BH statistical analysis, we selected 66 significant honeysuckle targets to perform enrichment analysis via CPDB. To begin, 15 pathways were selected from the top 150 pathways (ranked by p-value < 0.001) using the following keywords: virus, viral, infection, microRNA (miRNA), immune, inflammation, TNF, interleukin (IL), interferon (IFN), cytokine, etc. Those pathways highlighted in black were focused upon to discuss the possible therapeutic potential of TCMs. The 15 selected pathways in this study were illustrated in the context of a network. The dot color denotes statistical significance where the darker color demonstrates a higher significance level. The dot size indicates the number of genes in the pathway, while the edge between two dots showed the relationship. (B) The 64 target genes of Huangqi were analyzed though the same process in CPDB. The bar chart shows the results of the IPA analysis of honeysuckle (C) and Huangqi (D). Since the target genes in this study were retrieved from a database or PubMed, there were no up- or down-regulated expression data. Thus, we predicted the potential pathway without up- or down-regulation. A white bar color (no activation change) or gray (unknown activation) was the degree of correlation with the input gene set. The pathways of interest in this study were highlighted red.

FIGURE 3

Prediction of the functions of let-7a and miR-148b and their connections with the TCM candidates using multiple databases. miRNA-related targets were obtained from miRDB, and each list of genes (990 target genes for let-7a-5p and 499 target genes for miR-148b-5p) were used to query CPDB for signaling pathways. (A,B) show the network of 10 signaling pathways related to this study and mediated by let-7a-5p and miR-148b-5p, respectively. Pathways in black are highly related to this study. (C,D) display the intersecting target genes of miRNA and TCM candidates. Of particular interest, the MAPK pathway (yellow) and the IL6 pathway (red) are associated with cytokine storms in COVID-19 patients.

FIGURE 4

Induction of let-7a, miR-148b, and miR-146a levels by the TCM candidates. let-7a, miR-148b, and miR-146a levels were measured by qRT-PCR normalized to the internal control U54 after a 24-h treatment with honeysuckle-EtOH (A-C), honeysuckle-H2O (D-F), APS-L (G-I), and APS (J-L). Data are expressed as mean ± SD from three repeated results and analyzed using Student’s t-test. ^*, significantly different from the corresponding control, at p < 0.05; ^**, p < 0.01; ^***, p < 0.001.

FIGURE 5

Inhibition of cytokine release level by the TCM candidates. After 6 and 24 h of honeysuckle-EtOH (A) or APS-L (B) treatment in the presence or absence of LPS, THP-1 cell medium was collected, and ELISA was performed to measure cytokine levels. Treatment with LPS alone was considered to be the control. (C) After 24 h of honeysuckle-EtOH treatment in the presence or absence of stimulators (LPS, Spike, LPS + Spike, and R848), THP-1 cell medium was collected, and ELISA was performed to measure IL-6 levels. Data are expressed as mean ± SD from three repeated results and analyzed using Student’s t-test. ^* compared to LPS stimulation alone, ^# compared to combined stimulation of LPS and Spike, and ^$ compared to R848 stimulation alone. ^*, significantly different from the corresponding control, at p < 0.05; ^**, p < 0.01. Similar for ^# and ^$.

FIGURE 6

Honeysuckle and APS-L inhibit SARS-CoV2 M^pro activity. (A) Activity of SARS-CoV-2 M^pro in the presence of a peptide substrate, a fluorogenic probe (Abz-TSAVLQSGFRK-Dnp). The concentrations of the peptide substrate varied from 2 to 40 µM in PBS buffer, while M^pro concentration was fixed at 0.12 µM. The effect of honeysuckle-EtOH (B) and APS-L (C) on SARS-CoV-2 M^pro activity was investigated. The IC₅₀ value for each reaction was calculated and displayed.

FIGURE 7

The effect of honeysuckle, Huangqi, and their combination on the binding of SARS-CoV2-spike with ACE2 and syncytia formation. (A,C) Imaging of EGFP/spike-positive BHK-21 effector cell binding (upper panel) and fusion (lower panel) with Calu-3 target cells in the absence or presence of honeysuckle (A), Huangqi (C), or combination of honeysuckle-EtOH and either APS or APS-L. The binding efficiency of SARS-CoV2-spike to ACE2 (gray bars) and the formation of syncytium indicating fusion efficiency (white bars) was quantified in the cells treated with honeysuckle (B), Huangqi (D), and honeysuckle-EtOH combined with either APS or APS-L (E). *, p < 0.05; **, p < 0.01; ***, p < 0.001; n.s: no significance. Scale bar equals 1.0 mm in all figures.