Introduction
The coronavirus 2019 (COVID-19) pandemic was caused by a severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection.1 To date, although several vaccines have been used for the prevention of COVID-19, novel coronavirus variants have spread worldwide. In March 2021, the Delta SARS-CoV-2 variant was identified in India and has now spread to >100 countries. However, there is no clinically effective therapy against the SARS-CoV-2 infection. Of note, traditional Chinese medicine (TCM) has unique advantages in anti-inflammation and control of infectious diseases. The QingFeiPaiDu decoction (QFPDD) is composed of four classic prescriptions of ephedra, apricot kernel, gypsum, and licorice decoction, the Belamcanda and ephedra decoction, the Minor Bupleurum Decoction, and the Powder of Five Ingredients with Poria. A previous study showed that treatment with QFPDD effectively relieved fever, cough, fatigue, and other symptoms in newly diagnosed COVID-19 patients and rapid cleans SARS-CoV-2 to prevent severe disease as well as reduces the mortality rate of hospitalized severe COVID-19 patients.2 Currently, QFPDD has been approved for the treatment of COVID-19 patients by the Chinese government. Therefore, it is of great significance to explore the mechanisms that underlie the action of QFPDD in the treatment of COVID-19.
QFPDD contains 21 TCMs with complicated ingredients, and the clinical efficacy of QFPDD could result from combinational effects of its multi-ingredients that target several molecules and pathways as well as systemic body regulation. To develop new anti-SARS-CoV-2 drugs, a network pharmacological analysis3 was used to explore the active ingredients and potential molecular mechanisms that underlie the anti-inflammatory activity of QFPDD for the treatment of COVID-19, based on the main clinical symptoms of COVID-19 patients. The research workflow is shown in Figure 1.
Materials and methods
Collection of the active ingredients of QFPDD
Using the Traditional Chinese Medicine Systems Pharmacology (www.tcmspw.com/index.php ) and Traditional Chinese Medicines Integrated Database (www.119.3.41.228:8000/tcmid/search ), the potential active ingredients of QFPDD were screened, according to the criteria of oral bioavailability (OB) ≥30% and a drug-likeness (DL) ≥0.18. Gypsum fibrosum is a mineral medicine with a relatively simple composition. Gypsum fibrosum was included in the study although it did not meet the criteria.
Prediction of the targets of active ingredients of QFPDD
Using the PubChem database (www.pubchem.ncbi.nlm.nih.gov ) and the SwissTargetPrediction database (www.swisstargetprediction.ch ), the potential targets for the active ingredients of QFPDD were predicted (eliminated if p = 0 and the target was screened if p>median).
Screening of the main active ingredients and targets of QFPDD
The main active ingredients and targets for QFPDD were screened, according to the average degree of freedom and the maximum degree of freedom. The ingredient–target network diagram was constructed using Cytoscape software (version 3.7.2), and the network topology parameters were calculated using the Cytoscape tools.
Construction and analysis of the QFPDD-COVID-19-Target network
The main active ingredient targets and disease targets were established as pairs to obtain the potential targets for QFPDD in the treatment of COVID-19. The obtained data were used to construct the QFPDD-COVID-19-Target network using Cytoscape software (version 3.7.2). The topological parameters of the constructed network were analyzed to explore the interaction between different TCMs in QFPDD for the treatment of COVID-19.
Construction and analysis of potential protein–protein interactions
The potential targets of QFPDD in the treatment of COVID-19 were imported into the STRING database (www.string-db.org ), and the protein–protein interaction (PPI) network of potential targets was obtained, based on a confidence score of ≥0.9 for the interaction targets. Then, the topology parameters that included each node degree, the betweenness centrality (BC), closeness centrality (CC), average shortest path length (ASPL), and the average values of the potential targets for QFPDD were analyzed using Cytoscape software (version 3.7.2).
Gene Ontology enrichment analysis and Kyoko Encyclopedia of Genes and Genomes pathway analysis
The potential targets of QFPDD for the treatment of COVID-19 were analyzed by Gene Ontology (GO) enrichment analysis, and the potential human signal pathways of the targets involved were analyzed using the DAVID website (www.david.ncifcrf.gov ) and Kyoko Encyclopedia of Genes and Genomes (KEGG) database with a p-value of 0.05.
Molecular docking
The three-dimensional (3D) structures of the SARS-CoV-2 3-chymotrypsin-like cysteine protease (3CLpro) (PDB ID: 6LU7) and angiotensin-converting enzyme 2 (ACE2) (PDB ID: 1R42) were downloaded from the RCSB database (www.rcsb.org ). The original receptor proteins were extracted, and water molecules, other ligands, and residues were removed using PyMOL 1.7.2.1 software. AutoDock Tools 1.5.6 software was used for hydrogenation, merging of nonpolar and hydrogen interactions, and conversion into PDBQT format. The two-dimensional (2D) or 3D structures of the ligand compounds were downloaded from the PubChem database. The energy of the ligand compounds was minimized and saved in mol2 format with ChemBioOffice 2016 software. All of the flexible bonds of the ligand compounds were rotated by default. The grid box that was used for molecular docking included all the surrounding residues and centered on the primary ligands of the receptor proteins. The molecular docking was performed using Autodock Vina 1.1.2 and a Python script, and the sequences were sorted, according to the optimal affinity of each ligand compound.
Real-Time qPCR verifies the effect of QFPDD on gene expression
Gypsum fibrosum was decocted first, and all the TCMs were decocted twice with 1:8 (weight/weight) of water. The TCM liquid was concentrated to 76 mL, freeze-dried and ground into a fine powder with liquid nitrogen.
Human lung cancer A549 cells were obtained from the Chinese Academy of Sciences Stem Cell Bank. The cells (1 × 105 cells/well) were stimulated with 10 µg/mL lipopolysaccharides (LPS; L8880, Solarbio, Beijing, China) alone or LPS combined with 6.4 mg/mL QFPDD or 0.02 µmol/L Dexamethasone (DXM), for 24 h. The cells were harvested, and their total RNA was extracted, followed by reverse transcription into cDNA using a reverse transcription kit (RR047A, Takara, Beijing, China). The relative levels of the target to the control β-actin mRNA transcripts were quantified by Real-Time quantitative polymerase chain reaction (RT-qPCR) using specific primers (Table 1). The data were analyzed by 2−ΔΔCT.
Table 1Sequences of primers for RT-qPCR
| Target gene | Forward sequence and reverse sequence | Product length (bp) |
|---|
| β-actin | F: TGACGTGGACATCCGCAAAG | 204 |
| R: CTGGAAGGTGGACAGCGAGG | |
| SRC | F: GCTTCAACTCCTCGGACACC | 168 |
| R: ACCAGTCTCCCTCTGTGTTGTT | |
| MAPK1 | F: CCAGGGAAGCATTATCTTGACC | 172 |
| R: CTTTGGAGTCAGCATTTGGGAA | |
| PRKCD | F: AGCACAGAGCGTGGGAAAAC | 146 |
| R: GTCACCTCAGACACTGGCTCCT | |
| PTK2 | F: ATCTATCCAGGTCAGGCATCTCT | 179 |
| R: ATTTCCTGTTGCTGTCGGATTAG | |
| PRKCE | F: ATGCCCCCATAGGCTACGAC | 153 |
| R: CACCCGACGACCCTGAGAG | |
Statistical analysis
All data are expressed as mean ± standard deviation. Analysis of variance (ANOVA) was used to evaluate the statistical difference in data that met normal distribution and homogeneity of variance. A rank sum test was used when data did not meet a normal distribution or heterogeneity of variance. p>0.05 indicated no statistical difference. p<0.05 indicated statistical difference. p<0.01 indicated a significant difference.
Results
Active ingredients of QFPDD
There were 376 ingredients with an OB ≥30% and a DL ≥0.18 in the 20 TCMs and gypsum fibrosum: Ephedrae Herba (EH) contained 23, Glycyrrhizae Radix et Rhizoma (GRR) 92, Armeniacae by Amarum (ASA) 19, Gypsum Fibrosum (GF) 1, Cinnamomi Ramulus (CR) 7, Alismatis Rhizoma (AR) 10, Polyporus (POL) 11, Atractylodis Macrocephalae Rhizoma (AMR) 7, Poria (POR) 15, Bupleuri Radix (BUP) 17, Scutellariae Radix (SR) 36, Pinelliae Rhizoma (PR) 13, Zingiberis Rhizoma Recens (ZRR) 5, Asteris Radix et Rhizoma (AST) 19, Farfarae Flos (FF) 22, Belamcandae Rhizoma (BR) 17, Asari Radix et Rhizoma (ARR) 8, Dioscoreae Rhizoma (DR) 16, Aurantii Fructus Immaturus (AFI) 22, Citri Reticulatae Pericarpium (CRP) 5, and Pogostemonis Herba (PH) with 11 ingredients (Supplementary Document 1, sheet 1). Because Asteris Radix et Rhizoma and Asari Radix et Rhizoma have the same Latin initials, to distinguish between both TCMs, the first three letters of the first word of Asteris Radix et Rhizoma were used to represent this TCM.
The potential targets of active ingredients in QFPDD
A total of 18,833 targets were predicted by the 376 active ingredients in QFPDD (except for the targets of GF and some active ingredients that had not been identified). The detailed information on the targets of 20 TCMs’ active ingredients in QFPDD is shown in Supplementary Document 2.
Main active ingredients of QFPDD and their potential targets
Many ingredients of QFPDD might be helpful in the treatment of COVID-19; however, the main ingredients of QFPDD were used to elucidate the mechanism that underlies its pharmacological action. An ingredient–target network was constructed (Fig. 2) using these 20 TCMs and their abbreviations in Supplementary Document 1 (sheet 2). According to the ingredient–target network, there were 39 core ingredients and their corresponding targets (Table 2).
Table 2Main or core active ingredients and their targets in 20 TCMs
| Chinese name | Latin name | Number of main ingredients | Core ingredients | Core targets |
|---|
| 麻黄 | Ephedrae Herba | 16 | diosmetin (EH-10); genkwanin (EH-14); kaempferol (EH-04) | CYP19A1 |
| 炙甘草 | Glycyrrhizae Radix et Rhizoma | 75 | 3′-Methoxyglabridin (GRR-63); glypallichalcone (GRR-26); glyasperin B (GRR-18) | CYP19A1; PTPN1; ESR2; ESR1 |
| 杏仁 | Armeniacae Semen Amarum | 17 | gondoic acid (ASA-13); spinasterol (ASA-08) | ESR1; ESR2 |
| 桂枝 | Cinnamomi Ramulus | 3 | peroxyergosterol (CR-03); sitosterol (CR-02); beta-sitosterol (CR-01) | HSD11B2; CDC25A; NOS2; NR1H2; GLRA1; DHCR7; CDC25B |
| 泽泻 | Alismatis Rhizoma | 9 | alisol,b,23-acetate (AR-04); [(1S,3R)-1-[(2R)-3,3-dimethyloxiran-2-yl]-3-[(5R,8S,9S,10S,11S,14R)-11-hydroxy-4,4,8,10,14-pentamethyl-3-oxo-1,2,5,6,7,9,11,12,15,16-decahydrocyclopenta[a]phenanthren-17-yl]butyl] acetate (AR-08); alisol B monoacetate (AR-03) | AR; CES2; CYP17A1; HMGCR; HSD11B1; PTPN1 |
| 猪苓 | Polyporus | 8 | (22e,24r)-ergosta-6-en-3beta,5alpha,6beta-triol (POL-03); Cerevisterol (POL-01) | AR |
| 白术 | Atractylodis Macrocephalae Rhizoma | 3 | α-Amyrin (AMR-01); (3S,8S,9S,10R,13R,14S,17R)-10,13-dimethyl-17-[(2R,5S)-5-propan-2-yloctan-2-yl]-2,3,4,7,8,9,11,12,14,15,16,17-dodecahydro-1H-cyclopenta[a]phenanthren-3-ol (AMR-02) | CYP19A1; AR; CDC25A |
| 茯苓 | Poria | 14 | 3beta-Hydroxy-24-methylene-8-lanostene-21-oic acid (POR-09); trametenolic acid (POR-02); dehydroeburicoic acid (POR-14) | SHBG; PTPN1; HSD11B1 |
| 柴胡 | Bupleuri Radix | 13 | 3,5,6,7-tetramethoxy-2-(3,4,5-trimethoxyphenyl)chromone(BUP-07); quercetin (BUP-01); areapillin (BUP-08) | CYP19A1 |
| 黄芩 | Scutellariae Radix | 32 | 5,8,2′-Trihydroxy-7-methoxyflavone (SR-13); moslosooflavone (SR-28) | ESR2 |
| 姜半夏 | Pinelliae Rhizoma | 10 | beta-sitosterol (PR-01); gondoic acid (PR-08) | ACHE; CYP19A1; ESR2 |
| 生姜 | Zingiberis Rhizoma Recens | 4 | beta-sitosterol (ZRR-01); poriferast-5-en-3beta-ol (ZRR-03); stigmasterol (ZRR-02) | AR; ACHE; SLC6A2; BCHE; SLC6A4; CHRM2; CDC25A; CDC25B |
| 紫菀 | Asteris Radix et Rhizoma | 14 | quercetin (AST-02); galangin (AST-06) | ACHE; CYP19A1 |
| 冬花 | Farfarae Flos | 8 | quercetin (FF-01); femara (FF-08); tussilagolactone (FF-06) | CDK1 |
| 射干 | Belamcandae Rhizoma | 15 | epianhydrobelachinal (BR-10); anhydrobelachinal (BR-06) | CYP19A1; CA2 |
| 细辛 | Asari Radix et Rhizoma | 8 | 4,9-dimethoxy-1-vinyl-$b-carboline (ARR-07) | CDK1; CDK2; CYP19A1; CA12 |
| 山药 | Dioscoreae Rhizoma | 10 | denudatin B (DR-01); kadsurenone (DR-02) | CYP19A1; PTPN1; AR |
| 枳实 | Aurantii Fructus Immaturus | 20 | sinensetin (AFI-03) | ADORA1 |
| 陈皮 | Citri Reticulatae Pericarpium | 5 | 5,7-dihydroxy-2-(3-hydroxy-4-methoxyphenyl)chroman-4-one (CRP-03); naringenin (CRP-02) | ABCG2; ADORA1; ADORA3; CYP19A1; CYP1B1 |
| 藿香 | Pogostemonis Herba | 7 | quercetin (PH-01) | ADORA1 |
Targets are associated with clinical symptoms of COVID-19
To explore the molecular mechanism that underlies the action of QFPDD, 9,624 targets that were associated with common COVID-19 clinical symptoms, such as fever, cough, and fatigue were screened from the database (Supplementary Document 1, sheet 3).
QFPDD-COVID-19-Target network
Different TCMs might act synergistically on the same target when treating COVID-19 and enhance the therapeutic efficacy. A network of TCMs and the targets of ≥6 TCMs was generated in Figure 3. Among these targets, 19 TCMs targeted CYP19A1, and 18 targeted ESR2; 15 TCMs shared 3 common targets, and 14 targeted SHBG. In addition, 13 TCMS mutually targeted 4 targets, 12 TCMs shared 9 targets, 11 TCMs targeted 2 targets, 10 TCMs shared13 targets, 9 TCMs had the same 14 targets, 8 TCMs had 14 targets, 7 TCMs had 14 targets, and 6 TCMs shared14 targets. Therefore, TCMs had a multi-ingredient-multitarget feature for the treatment of COVID-19.
PPI network
The potential targets of QFPDD for the treatment of COVID-19 were analyzed using the STRING database, and the PPI network of protein interactions was obtained. Through the calculation of the degree of each node, the BC, CC, and the ASPL, 24 key targets were selected (Table 3).
Table 3Topological analysis of the QFPDD-COVID-19 Target network
| Target name | Abbreviation | Uniprot ID | ASPL | BC | CC | Degree |
|---|
| Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha | PIK3CA | P42336 | 2.189 | 0.098 | 0.457 | 66 |
| Phosphoinositide-3-kinase regulatory subunit 1 | PIK3R1 | P27986 | 2.205 | 0.086 | 0.454 | 62 |
| Amyloid beta precursor protein | APP | P05067 | 2.358 | 0.166 | 0.424 | 55 |
| SRC proto-oncogene, non-receptor tyrosine kinase | SRC | P12931 | 2.280 | 0.053 | 0.439 | 49 |
| mitogen-activated protein kinase 1 | MAPK1 | P28482 | 2.130 | 0.084 | 0.470 | 47 |
| mitogen-activated protein kinase 3 | MAPK3 | P27361 | 2.224 | 0.042 | 0.450 | 43 |
| AKT serine/threonine kinase 1 | AKT1 | P31749 | 2.268 | 0.063 | 0.441 | 40 |
| heat shock protein 90 alpha family class A member 1 | HSP90AA1 | P07900 | 2.311 | 0.065 | 0.433 | 38 |
| E1A binding protein p300 | EP300 | Q09472 | 2.453 | 0.065 | 0.408 | 34 |
| cyclin dependent kinase 1 | CDK1 | P06493 | 2.614 | 0.033 | 0.382 | 30 |
| Janus kinase 2 | JAK2 | O60674 | 2.512 | 0.017 | 0.398 | 30 |
| epidermal growth factor receptor | EGFR | P00533 | 2.370 | 0.041 | 0.422 | 29 |
| Thrombin | F2 | P00734 | 2.508 | 0.033 | 0.399 | 28 |
| mitogen-activated protein kinase 8 | MAPK8 | P45983 | 2.283 | 0.045 | 0.438 | 28 |
| retinoid X receptor alpha | RXRA | P19793 | 2.500 | 0.072 | 0.400 | 27 |
| estrogen receptor 1 | ESR1 | P03372 | 2.382 | 0.090 | 0.420 | 27 |
| protein kinase C delta | PRKCD | Q05655 | 2.563 | 0.015 | 0.390 | 23 |
| protein tyrosine kinase 2 | PTK2 | Q05397 | 2.618 | 0.012 | 0.382 | 22 |
| ribosomal protein S6 kinase B1 | RPS6KB1 | P23443 | 2.480 | 0.012 | 0.403 | 21 |
| mitogen-activated protein kinase 14 | MAPK14 | Q16539 | 2.567 | 0.012 | 0.390 | 21 |
| nuclear receptor subfamily 3 group C member 1 | NR3C1 | P04150 | 2.430 | 0.022 | 0.412 | 21 |
| protein kinase C epsilon | PRKCE | Q02156 | 2.594 | 0.015 | 0.385 | 17 |
| cyclin dependent kinase 5 | CDK5 | Q00535 | 2.634 | 0.028 | 0.380 | 17 |
| peroxisome proliferator activated receptor alpha | PPARA | Q07869 | 2.587 | 0.031 | 0.387 | 15 |
GO enrichment analysis and KEGG pathway analysis
The GO enrichment analysis of the potential targets of QFPDD for the treatment of COVID-19 revealed 574 GO entries with p<0.05, a minimum count of 3 and an enrichment factor >1.5 (the enrichment factor is the ratio between the observed number and the number expected by chance), which included 353 entries for biological processes, 112 for molecular function, and 109 for cell composition. The top enriched 20 biological process entries in each category are shown in Figure 4a. The KEGG pathway analysis indicated that 214 signal pathways were involved. They included the PI3K-Akt, JAK-STAT, AMPK, NOD-like receptor, NF-κB, Notch, TGF-β, and HIF-1 signal pathways4–15 that are associated with the pathogenesis of pneumonia. The top 20 signal pathways are shown in Figure 4b. There were 196 signal pathways with a p<0.01, and 262 genes were associated with these pathways. Their correlation was established in Figure 5. The 24 critical targets that were identified by network analysis had high degree values. These genes might be the critical target genes of QFPDD for the treatment of COVID-19. According to the statistics on the occurrence of the 24 critical targets in 196 signaling pathways, 24 targets were involved in 156 signal pathways (Fig. 6), which included pathways involved in cancer, endocrine resistance, the PI3K-Akt signal pathway, proteoglycans in cancer, and the Rap1 signal pathway. The Rap1 signaling pathway is crucial for the pathogenesis of acute lung injury or acute respiratory distress syndrome (ALI/ARDS), and the regulation of the Rap1 signaling pathway might become a new strategy for the treatment of ALI/ARDS.16 A diagram of the 24 critical targets for the treatment of COVID-19 is shown in Figure 7.
Molecular docking results
In total, 375 active ingredients from 20 TCMs (except for GF) in QFPDD were selected and screened for their binding to ACE2 and 2019-nCoV 3CLpro by molecular docking, based on the binding energy. The lower the binding energy, the higher the possibility of binding. The docking results indicated that the binding energies of the 375 main active ingredients were all <0 (except for taraxanthin and delphinidin). The 39 core ingredients screened by the ingredient–target network under 3.3 displayed a good binding affinity to the 3CLpro and ACE2 proteins, and their binding energies were <–3.9. The docking results for each core ingredient for 3CLpro and ACE2 are given in Table 4. To further analyze the interaction between the ingredient and the protein, 2D and 3D molecular docking diagrams exhibited the potential interactions between the ingredient and 3CLpro or ACE2. The molecular docking of the ingredients with the best results is shown in Figure 8. Due to the presence of a hydrophobic Pi-alkyl, the 3′-methoxyglabridin parent nucleus was mainly bound to the amino acid residues CYS145, MET165, and PRO168 of 3CLpro. In addition, hydrogen bonds formed between the phenolic hydroxyl groups on the parent nucleus and the amino acid residues GLY143, LEU141, and CYS145. Due to the presence of the hydrophobic Pi-alkyl, the Alisol B 23-acetate parent nucleus bound to the ACE2 amino acid residues LEU85 and HIS15, and the hydroxyl and carbonyl groups on the parent nucleus formed hydrogen bonds with the amino acid residues GLN101 and ASN103. These interactions might increase the binding of the molecules to the proteins.
Table 4Docking of the 39 core components with 3CLpro and ACE2
| Number | Compound | Binding energy values with 3CLpro (kcal/mol) | Binding energy values with ACE2 (kcal/mol) |
|---|
| 1 | 3′-Methoxyglabridin | −8.3 | −6.1 |
| 2 | Moslosooflavone | −7.8 | −6.0 |
| 3 | Kaempferol | −7.7 | −5.9 |
| 4 | Alisol,b,23-acetate | −7.7 | −6.8 |
| 5 | Quercetin | −7.6 | −6.2 |
| 6 | Peroxyergosterol | −7.6 | −5.8 |
| 7 | Diosmetin | −7.6 | −6.3 |
| 8 | Genkwanin | −7.6 | −5.9 |
| 9 | Epianhydrobelachinal | −7.6 | −6.5 |
| 10 | α-Amyrin | −7.5 | −6.7 |
| 11 | [(1S,3R)-1-[(2R)-3,3-dimethyloxiran-2-yl]-3-[(5R,8S,9S,10S,11S,14R)-11-hydroxy-4,4,8,10,14-pentamethyl-3-oxo-1,2,5,6,7,9,11,12,15,16-decahydrocyclopenta[a]phenanthren-17-yl]butyl] acetate | −7.5 | −6.8 |
| 12 | Areapillin | −7.4 | −6.3 |
| 13 | Naringenin | −7.3 | −5.9 |
| 14 | Glyasperin B | −7.3 | −5.7 |
| 15 | 5,8,2′-Trihydroxy-7-methoxyflavone | −7.2 | −6.1 |
| 16 | Galangin | −7.2 | −6.0 |
| 17 | Spinasterol | −7.1 | −5.6 |
| 18 | Cerevisterol | −7.1 | −5.6 |
| 19 | 3beta-Hydroxy-24-methylene-8-lanostene-21-oic acid | −7.0 | −6.3 |
| 20 | Poriferast-5-en-3beta-ol | −7.0 | −5.5 |
| 21 | Stigmasterol | −7.0 | −5.5 |
| 22 | beta-sitosterol | −6.9 | −6.8 |
| 23 | Sinensetin | −6.9 | −5.5 |
| 24 | Dehydroeburicoic acid | −6.8 | −5.8 |
| 25 | Alisol B monoacetate | −6.8 | −6.0 |
| 26 | (22e,24r)-ergosta-6-en-3beta,5alpha,6beta-triol | −6.8 | −5.5 |
| 27 | Glypallichalcone | −6.7 | −5.4 |
| 28 | Femara | −6.7 | −5.3 |
| 29 | 5,7-dihydroxy-2-(3-hydroxy-4-methoxyphenyl)chroman-4-one | −6.6 | −5.5 |
| 30 | Trametenolic acid | −6.6 | −6.4 |
| 31 | 3,5,6,7-tetramethoxy-2-(3,4,5-trimethoxyphenyl)chromone | −6.5 | −5.4 |
| 32 | Sitosterol | −6.5 | −5.7 |
| 33 | Tussilagolactone | −6.5 | −4.8 |
| 34 | Denudatin B | −6.5 | −5.4 |
| 35 | Anhydrobelachinal | −6.5 | −5.8 |
| 36 | (3S,8S,9S,10R,13R,14S,17R)-10,13-dimethyl-17-[(2R,5S)-5-propan-2-yloctan-2-yl]-2,3,4,7,8,9,11,12,14,15,16,17-dodecahydro-1H-cyclopenta[a]phenanthren-3-ol | −6.4 | −5.0 |
| 37 | 4,9-dimethoxy-1-vinyl-$b-carboline | −6.3 | −5.7 |
| 38 | Kadsurenone | −6.1 | −5.2 |
| 39 | Gondoic acid | −4.7 | −3.9 |
The Effect of QFPDD on the levels of mRNA transcripts of SRC, MAPK1, PRKCD, PTK2, and PRKCE
In total, 5 out of 24 key target genes with different degree values were selected to verify the effect of QFPDD on inflammation. As shown in Figure 9, compared with the control group, the relative levels of SRC, MAPK1, PRKCD, PTK2, and PRKCE transcripts increased significantly, which were significantly mitigated in the QFPDD and DXM groups. Of note, the relative levels of SRC and MAPK1 mRNA transcripts in the QFPDD group were slightly higher than those in the DXM group, although they were statistically insignificant.
Discussion
Network pharmacology is based on the systematic interactions between drugs, targets, and diseases and uses complex network models to determine the pharmacological properties of the research objects.17 The research and development of network pharmacology in TCMs provided a new perspective for the systematic study of the complex ingredients of TCMs.18–20 Therefore, in this study, the target network for QFPDD and COVID-19 was constructed and analyzed, and the targets were used for functional enrichment and signal pathway analysis to reveal the potential molecular mechanisms that underlie the action of QFPDD for the treatment of COVID-19 and to identify the main active ingredients and potential target genes of QFPDD.
Main active ingredients
Previous studies confirmed that SARS-CoV-2, similar to SARS-CoV, binds to ACE2 and causes infection.21 Therefore, the ingredient most closely related to the hydrolase and ACE2 of 2019-nCoV 3CLpro might be the main bioactive ingredient of QFPDD against SARS-CoV-2-induced COVID-19. This study used 20 TCMs (except for GF) from QFPDD to screen 375 major active ingredients and tested their binding to 3CLpro and ACE2 by molecular docking. The results indicated that except for taraxanthin and delphinidin, the remaining ingredients had binding energies <0 when docking with both proteins, which suggested that these ingredients from QFPDD might bind directly to the proteins, which ameliorated SARS-CoV-2 infection-induced COVID-19. The 39 core ingredients identified through network pharmacology had many targets that might improve disease symptoms and could be closely associated with 3CLpro and ACE2. Therefore, the 39 ingredients might be the main bioactive ingredients for the treatment of COVID-19 by QFPDD.
Potential target genes
This study found 24 key targets (PIK3CA, PIK3R1, APP, SRC, MAPK1, MAPK3, AKT1, HSP90AA1, EP300, CDK1, JAK2, EGFR, F2, MAPK8, RXRA, ESR1, PRKCD, PTK2, RPS6KB1, MAPK14, NR3C1, PRKCE, CDK5, and PPARA) for the treatment of COVID-19. Due to their close association, these targets might interact and act synergistically for the pathogenesis of COVID-19. Aberrant SRC activity contributes to the pathogenesis of pneumonia, which represents a potential target for pneumonia treatment.22 MAPK activation modulates the inflammatory response in ALI. MAPK1 activation reduces the LPS-induced inflammatory injury in A549 cells and ameliorates lung injury in ALI mice.23 PRKCD is an important regulator of human neutrophil proinflammatory responses and is crucial to regulate neutrophil–endothelial cell interactions and recruitment in inflamed lungs.24 The activation of members of the PTK family participate in acute inflammatory responses, during ALI and sepsis, and are essential for the recruitment and activation of monocytes, macrophages, neutrophils, and other immune cells.25 PRKCE could be used as a breathing regulator by affecting mitochondrial respiration, which promotes the occurrence of mitochondrial organisms and reduces organ damage that is caused by pneumonia.26 COVID-19 causes severe inflammation and acute lung damage, which leads to severe respiratory distress syndrome. In this study, a cellular model of inflammation was used to verify the relevant targets for QFPDD for inflammation. We found that treatment with QFPDD significantly mitigated the LPS-induced gene expression in A549 cells and that these genes belonged to the 24 potential targets of QFPDD. These findings might help in understanding QFPDD treatment for COVID-19.
Functional enrichment and pathway analysis
The GO functional enrichment analysis revealed the potential targets that were involved in biological processes, molecular function, and cell components.27 The potential targets of QFPDD were enriched in biological processes that included the cellular response to nitrogen compounds and the positive regulation of transferase activity. A previous study showed that the inhibition of nitric oxide (NO) led to an increase in the release of proinflammatory cytokines, deteriorating lung injury, which suggested that NO might ameliorate ALI.28 In addition, QFPDD targeted molecules that were mainly associated with protein kinase activity, protein tyrosine kinase activity, and kinase binding. Experimental results have shown that the mitochondrial antioxidant pathway mediated by protein kinase D1 is crucial for the early stage of bleomycin-induced pulmonary fibrosis in rats.29 Similarly, protein kinase CK2 is related to the prognosis of patients with lung cancer, which suggests that protein kinase CK2 might be a potential target for the treatment of lung cancer.30 In addition, QFPDD targeted cell components that are mainly associated with the membrane, receptor complex, and dendrite. Experiments demonstrated that the lipid raft protein stomatin, in the presence of low oxygen and Dex upregulation stabilized the cytoskeleton that is connected to the membrane and increased the barrier function of lung epithelial cells, which protected lung tissue cells.31 The targets by QFPDD might improve the symptoms of COVID-19 patients through the cellular response to nitrogen compounds, protein kinase activity, and membrane rafts. Therefore, the treatment of COVID-19 by QFPDD might be more closely related to these biological processes.
The KEGG pathway analysis exhibited that the enriched pathways were involved in cancer, the neuroactive ligand-receptor interaction, and the PI3K-Akt signal pathway. The 24 predicted key targets of QFPDD were more closely related to the pathways involved in cancer and proteoglycans in cancer, which suggested that the pathogenesis of COVID-19 might have similar processes to cancer. A recent study demonstrated that the risk of COVID-19 infection in cancer patients was 2.31-fold higher than that of the general population in Wuhan.32 Cancer patients might have a low immune function, which could promote virus infection, or have a poor function in the key target genes, or both. Most of the 24 potential targets of QFPDD were closely related to cancer development.33 carcinogenic signal pathways, which included the PI3K-Akt-mTOR, MAPK-ERK, SRC, CDK4-CDK6, and ER pathways, which are critical for endocrine resistance. Targeting these pathways, especially the ER pathway, could be the most effective way to combat resistance to anti-estrogens, and clinical trials have shown promising results. The PI3K-Akt signal pathway regulates the cellular inflammatory response by stimulating the expression of endothelial nitric oxide synthase (eNOS) and increasing the production of NO to increase vascular permeability, promote the infiltration of inflammatory cells, stimulate the activation of inflammatory cells, and induce the secretion of a large number of inflammatory factors through the activity of NO.34 Studies demonstrated that PI3K inhibitors had a significant inhibitory effect on NO production and inflammatory response in lung tissues of a viral pneumonia mouse model induced by influenza A virus infection.35 In addition to TCMs for COVID-19, some anti-virus (HIV or Ebola) drugs inhibit SARS-CoV-2 replication to ameliorate COVID-19 in the clinic. In this study, we found that 24 genes participated in the signal pathways associated with human immunodeficiency virus infection, which might explain why remdesivir and lopinavir or ritonavir have some benefits to COVID-19 patients.
Future directions
QFPDD is characterized by multiple components, multiple targets, and multiple pathways to treat COVID-19, which could improve the clinical symptoms of patients. The 39 active components, 24 potential key targets, and signaling pathways, such as the PI3K-Akt signaling pathway in QFPDD we found could provide a basis for further study of QFPDD and provide a reference for the development of new drugs for the treatment of COVID-19. In the future, the 39 active ingredients and related signaling pathways should be verified to clarify the material basis and other pharmacological effects of QFPDD for the treatment of COVID-19, such as organ protection, enhancement of immunity, and other functions.