Abstract
As a popular Chinese herbal medicine (CHM), polygonum cuspidatum is widely used to treat various diseases in China. However, its biological function and action mechanism have yet to be systematically explored. In the present study, we first identified 14 potential active ingredients of polygonum cuspidatum using the TCMSP server and then conducted an in silico target prediction for these ingredients using PharmMapper. The subsequent KEGG pathway enrichment analysis of the 57 identified potential targets revealed that they were closely associated with cancer and gynecological disorders. Furthermore, a protein-protein interaction network of these targets was constructed using STRING and Cytoscape, through which 11 core targets were excavated according to degree, a key topological parameter. Meanwhile, we developed a novel formula, in which the “R value” is determined by average shortest path length and closeness centrality, two other key topological parameters, to evaluate the reliability of these predicted core targets. Intriguingly, among the top 10 core targets excavated using this new formula, 7 overlapped with the former 11 core targets, showing a good consistency in these core targets between the different prediction algorithms. Next, 7 ingredients were identified/validated from the crude extract of polygonum cuspidatum using UPLC-MS/MS. Noteworthy, 6 potential targets predicted for these 7 ingredients overlapped with the 7 core targets excavated from the previous in silico analyses. Further molecular docking and druggability analyses suggested that polydatin may play a pivotal role in manifesting the therapeutic effects of polygonum cuspidatum. Finally, we carried out a series of cell functional assays, which validated the anti-proliferative effects of polygonum cuspidatum on gynecological cancer cells, thus demonstrating our network pharmacology approach is reliable and powerful enough to guide the CHM mechanism study.
Keywords: Polygonum cuspidatum (Hu Zhang), Chinese herb medicine, network pharmacology, UPLC-MS/MS, polydatin
Introduction
From ancient times to the present, Chinese herbal medicine (CHM) has been applied to prevent and treat various diseases in China. Nowadays, owing to its distinct therapeutic effects, CHM has been attracting more and more attention worldwide [1-3]. Polygonum cuspidatum (Hu Zhang), widely distributed in southern China, is a popular CHM with a wide range of pharmacological and biological activities. To date, > 100 prescriptions containing this plant have been used to treat a large variety of diseases, such as infection, inflammation, skin burns and hyperlipemia [4,5]. However, current knowledge about the medicinal characteristics of polygonum cuspidatum is incomplete, and a deeper understanding of its system-level action mechanism is required.
Composed of multiple ingredients, CHM has holistic and synergistic effects on complicated diseases. However, the action mechanism study of CHM is hindered by the complicated interactions among its “multi-ingredients” and the affected “multi-targets”. Network pharmacology has been lately applied to CHM research and is considered to be a promising approach to overcoming the above obstacle. Based on the network of “drug-target-disease-pathway” interactions, network pharmacology can reveal the synergistic effects on human systems by complex CHM. Meanwhile, with the recent advances in computational methods, bioinformatics has been used for the target prediction for CHM, which has significantly promoted the mechanism study of CHM [6,7]. Thus, in conjunction with bioinformatics, network pharmacology is now a powerful tool for untangling the disease-disease, disease-target protein, target protein-drug and drug-drug interactions among multi-ingredient CHM and complicated diseases [8-11].
Herein, in order to unravel the action mechanism of polygonum cuspidatum, we first screened active ingredients in polygonum cuspidatum using the TCMSP server and a literature search, followed with ingredient-target prediction by reverse docking analysis. GO functional and KEGG pathway enrichment analyses of the potential protein targets were then performed and the following constructed drug-target-pathway network revealed a great diversity of biological functions of polygonum cuspidatum. Thus, by taking advantage of the recent advances in network pharmacology and bioinformatics, we provided an overview on the action mechanism of active ingredients in polygonum cuspidatum, achieving a comprehensive understanding of its holistic therapeutic effect on specific diseases.
Materials and methods
Cell culture and reagents
Human breast cancer MDA-MB-231 cells and ovarian cancer SKOV-3 cells were obtained from the China Infrastructure of Cell Line Resources (School of Basic Medicine Peking Union Medical College, China) and were maintained in RPMI-1640 medium supplemented with 10% (v/v) FBS and 100 U/ml streptomycin/penicillin in the humidified atmosphere of 5% CO2 at 37°C. RPMI-1640 medium were purchased from GIBCO (USA). Fetal bovine serum (FBS) was purchased from ThermoFisher (USA). Cell Counting Kit-8 (CCK-8) reagent was bought from Dojindo (Japan). Annexin-V FITC apoptosis detection kit (#556547) was purchased from BD Biosciences Pharmingen (USA). The standard of polygonum cuspidatum was obtained from China National Institutes for Food and Drug Control.
Screening of active ingredients and relative target prediction
The value of oral bioavailability (OB) was evaluated by the OBioavail 1.1 model in the TCMSP database (http://ibts.hkbu.edu.hk/LSPtcmsp.php). Meanwhile, drug-likeness (DL) was analyzed using a model from the TCMSP database, which was constructed on account of the molecular descriptors and Tanimoto coefficient [12,13]. The parameters for the selection of active ingredients were set as follows: OB ≥ 30% and DL ≥ 0.18. The Tripos mol2 type files of the selected active ingredients were searched using the TCMSP database. Next, the potential protein targets were predicted by importing each component file with mol2 format into the PharmMapper reverse docking database in May 2018. By normalized fit score, the top 10 high-matching targets for each of the active ingredients were selected and the overlapped ones were excluded. The selected targets were then searched in the Uniprot database to identify their human-related protein target codes.
GO and KEGG enrichment analysis, and systematic network construction
Functional enrichment of GO was performed using OmicShare tools, a free online platform for data analysis (http://www.omicshare.com/tools) [14]. Meanwhile, potential protein targets were uploaded to the DAVID Bioinformatics Resources 6.8 server (https://david.ncifcrf.gov/summary.jsp) for KEGG pathway enrichment analysis. In order to obtain a deeper understanding of the complex relationships among drug, protein targets and pathways, a systematic network was constructed using Cytoscape 3.2.1. In GO and KEGG analysis, P < 0.05 was defined as significantly enriched.
PPI network construction and excavation of core genes
All potential protein targets were uploaded to the STRING database (http://string-db.org/) to analyze their interactions. Comprehensive information on the interactions of these protein targets was downloaded from STRING, and the interactions with a combined score of > 0.4 were selected for the construction of the PPI network using Cytoscape 3.2.1. The core genes from the PPI network were excavated using Degree, a key topological parameter, with the cut-off criteria of DC scores of ≥ 10. Furthermore, the average shortest path length and closeness centrality, another two topological parameters, were calculated by NetworkAnalyzer [15]. R-value was used to determine the ranks of the targets of polygonum cuspidatum using the following formula:
 |
This formula used the idea of normalization, where xa was the average shortest path length, xb was the closeness centrality, and R was an indicator to evaluate the importance of a target.
Cellular functional assays
The cell viability was assessed by the CCK-8 assay, in which 10 μl of CCK-8 was added to each well, and the cells were incubated for an additional two hours before the absorbance value was measured at a wavelength of 450 nm using an automated microplate reader. Apoptosis detection and Western Blot assay were performed according to the manufacturer’s instructions.
Liquid chromatography, mass spectrometric conditions and sample extraction preparation
Liquid chromatographic separation and mass spectrometric detection were performed using Waters Quattro Premier XE/Acquity UPLC system coupled to a tandem quadrupole mass spectrometer (Waters Corporation, Milford, MA, USA). Lab solutions LCMS software (Masslynx V4.1) was used to control the instruments and process the data. This instrument was equipped with both ESI and APCI sources. Chromatographic separation was performed using an Agilent Eclipse PlusC18 column (100 × 4.6 mm, 3.5 µm) at a flow rate of 0.4 ml/min in ESI source. The isocratic mobile phase consisted of methanol and 0.01% formic acid with gradient elution; run times were from 0-20 min with 10-90% B, 20-22 min with 90% B, 22-25 min with 90-10% B and 25-30 min with 10% B. The mass spectrometer operating parameters were optimized as follows: Interface voltage, 3.2 kV; desolvation temperature, 450°C; source temperature, 120°C. Analyst software version Masslynx V4.1 was used to control all parameters of liquid chromatography and mass spectrometry (MS).
Total RNA extraction and quantitative real-time PCR
Total RNA from breast cancer and ovarian cancer cells were extracted using the TRIzol reagent. The total RNA was quantified using Nano-drop 1000 (Thermo Fisher Scientific, USA). Two microgram of the total RNA from each sample was used for cDNA synthesis with PrimeScript RT Master Mix reagent according to the manufacturer’s protocol. By using the specific primer pairs (Table 1) and SYBR GREEN Premix reagent, qRT-PCR was performed on the QuantStudio 5 real-time PCR system (Thermo Fisher Scientific, USA). Expression data were normalized to the geometric mean of the housekeeping GAPDH gene to control the variability in mRNA expression levels.
Table 1.
Primer sequences of mRNA structural components used in q-PCR
| Components | Position | Sequence |
|---|---|---|
| EGFR | Forward | AGGCACGAGTAACAAGCTCAC |
| Reverse | ATGAGGACATAACCAGCCACC | |
| CASP3 | Forward | CATGGAAGCGAATCAATGGACT |
| Reverse | CTGTACCAGACCGAGATGTCA | |
| ALB | Forward | TCGGCTTATTCCAGGGGTGT |
| Reverse | AAGGCAATCAACACCAAGGCT | |
| MAPK14 | Forward | GCACTGAAGAAGCTGTCGAG |
| Reverse | GAACGTGGTCATCGGTAAGC | |
| F2 | Forward | CTGAGGGTCTGGGTACGAACT |
| Reverse | TGGGTAGCGACTCCTCCATAG | |
| ESR1 | Forward | GGGAAGTATGGCTATGGAATCTG |
| Reverse | TGGCTGGACACATATAGTCGTT | |
| GAPDH | Forward | TGCACCACCAACTGCTTAGC |
| Reverse | GGCATGGACTGTGGTCATGAG |
Systemsdock analysis
SystemsDock online tools were first utilized to check the docking efficiency between the 6 core genes and 7 key ingredients. The PDB ID of 6 core targets and Tripos mol2 type files of these ingredients was inputted simultaneously into the system for further analysis. Next polydatin and CASP3 (PDB ID: 4QUJ) were re-docked by Autodock software, and the top 30 docking conformations were extracted and selected as the best phase of docking energy for further analysis.
Statistical analysis
All data were analyzed using SPSS17.0 software (SPSS, Inc., Chicago, IL, USA). Results are represented as the mean ± standard deviation. Student’s t-test is used for comparisons between different groups. P < 0.05 was considered to indicate a statistically significant difference.
Results
Selection of the potential active ingredients in polygonum cuspidatum and prediction of their potential protein targets
First, using the TCMSP server, a unique systems-level pharmacology platform for CHM, we calculated ADME-related characteristics for all ingredients of polygonum cuspidatum. The specific pharmacokinetic properties of these ingredients undergoing evaluation included OB, DL, CaCo-2 permeability, blood-brain barrier permeability and Lipinski’s rule of five (MW, AlogP, TPSA, Hdon and Hacc). Since OB and DL represent the most important pharmacokinetic properties, the selection parameter for the active ingredients was set as follows: OB ≥ 30% and DL ≥ 0.18 [12,13,16]. Based on this threshold, we obtained 9/62 active ingredients of polygonum cuspidatum, namely physovenine, rhein, catechin, quercetin, beta-sitosterol, picralinal, physciondiglucoside, torachrysone-8-O-beta-D-(6’-oxayl)-glucoside and luteolin. Another 5 key active ingredients, namely emodin, chrysophanol, resveratrol, polydatin and apigenin were also included for further analysis referring to previous studies [17,18], although they did not meet all aforementioned criteria. We therefore finally screened out 14 active ingredients of polygonum cuspidatum for further protein target prediction. The structure and ADME-related characteristics of the above ingredients are shown in Figure 1 and Table 2, respectively.
Figure 1.
Chemical structure of fourteen active ingredients of polygonum cuspidatum downloaded from the TCMSP database. These fourteen active ingredients are: physovenine (MOL013287), rhein (MOL002268), catechin (MOL000492), quercetin (MOL000098), beta-sitosterol (MOL000359), picralinal (MOL013288), physciondiglucoside (MOL002259), torachrysone-8-O-beta-D-(6’-oxayl)-glucoside (MOL002280), luteolin (MOL000006), emodin (MOL000472), chrysophanol (MOL001729), resveratrol (MOL012744), polydatin (MOL013289) and apigenin (MOL000008).
Table 2.
Pharmacological and molecular parameters of Polygonum cuspidatum
| MOL_ID | Name | MW | AlogP | Hdon | Hacc | OB (%) | CaCo-2 | BBB | DL | TPSA | RBN |
|---|---|---|---|---|---|---|---|---|---|---|---|
| MOL013287 | Physovenine | 262.34 | 2.08 | 1 | 5 | 106.21 | 0.51 | 0.20 | 0.19 | 50.80 | 2 |
| MOL002268 | Rhein | 284.23 | 1.88 | 3 | 6 | 47.07 | -0.20 | -0.99 | 0.28 | 111.90 | 1 |
| MOL000492 | Catechin | 290.29 | 1.92 | 5 | 6 | 54.83 | -0.03 | -0.73 | 0.24 | 110.38 | 1 |
| MOL000098 | Quercetin | 302.25 | 1.50 | 5 | 7 | 46.43 | 0.05 | -0.77 | 0.28 | 131.36 | 1 |
| MOL000359 | Beta-sitosterol | 414.79 | 8.08 | 1 | 1 | 36.91 | 1.32 | 0.87 | 0.75 | 20.23 | 6 |
| MOL013288 | Picralinal | 366.45 | 1.80 | 1 | 6 | 58.01 | 0.23 | -0.21 | 0.75 | 67.87 | 3 |
| MOL002259 | Physciondiglucoside | 608.60 | -0.91 | 8 | 15 | 41.65 | -2.64 | -3.43 | 0.63 | 242.13 | 7 |
| MOL002280 | Torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | 480.46 | 0.64 | 5 | 12 | 43.02 | -1.23 | -1.84 | 0.74 | 189.28 | 8 |
| MOL000006 | Luteolin | 286.25 | 2.07 | 4 | 6 | 36.16 | 0.19 | -0.84 | 0.25 | 111.13 | 1 |
| MOL000472 | Emodin | 270.25 | 2.49 | 3 | 5 | 24.40 | 0.22 | -0.66 | 0.24 | 94.83 | 0 |
| MOL001729 | Chrysophanol | 254.25 | 2.76 | 2 | 4 | 18.64 | 0.62 | -0.20 | 0.21 | 74.60 | 0 |
| MOL012744 | Resveratrol | 228.26 | 3.01 | 3 | 3 | 19.07 | 0.80 | -0.01 | 0.11 | 60.69 | 2 |
| MOL013289 | Polydatin | 390.42 | 1.11 | 6 | 8 | 21.44 | -0.90 | -1.81 | 0.50 | 139.84 | 5 |
| MOL000008 | Apigenin | 270.25 | 2.33 | 3 | 5 | 23.06 | 0.43 | -0.61 | 0.21 | 90.90 | 1 |
Next, we predicted the potential targets for each of these 14 ingredients using the PharmMapper reverse docking database. By normalized fit score, the top 10 high-matching targets for each of the 14 ingredients were selected for further analysis. After ruling out the ones that overlapped, we finally obtained a total of 57 potential protein targets for these 14 ingredients (Table 3).
Table 3.
The potential protein targets of 14 active ingredients in Polygonum cuspidatum
| No. | Compound | Protein code | Protein name | Normalized Fit Score |
|---|---|---|---|---|
| 1 | physovenine | P12643 | BMP2_HUMAN | 1 |
| 2 | physovenine | P00918 | CAH2_HUMAN | 0.9951 |
| 3 | physovenine | P09211 | GSTP1_HUMAN | 0.9721 |
| 4 | physovenine | P20701 | ITAL_HUMAN | 0.9492 |
| 5 | physovenine | P62937 | PPIA_HUMAN | 0.947 |
| 6 | physovenine | P06276 | CHLE_HUMAN | 0.9338 |
| 7 | physovenine | P52895 | AK1C2_HUMAN | 0.9275 |
| 8 | physovenine | P15121 | ALDR_HUMAN | 0.9 |
| 9 | physovenine | Q8NBQ5 | DHB11_HUMAN | 0.8863 |
| 10 | physovenine | P23946 | CMA1_HUMAN | 0.8543 |
| 11 | rhein | P00751 | CFAB_HUMAN | 0.9897 |
| 12 | rhein | P45983 | MK08_HUMAN | 0.9685 |
| 13 | rhein | P53779 | MK10_HUMAN | 0.9645 |
| 14 | rhein | Q16539 | MK14_HUMAN | 0.9632 |
| 15 | rhein | P08254 | MMP3_HUMAN | 0.9579 |
| 16 | rhein | P62937 | PPIA_HUMAN | 0.9431 |
| 17 | rhein | P11511 | CP19A_HUMAN | 0.9312 |
| 18 | rhein | O14965 | AURKA_HUMAN | 0.9268 |
| 19 | rhein | O14757 | CHK1_HUMAN | 0.9137 |
| 20 | rhein | Q8NBQ5 | DHB11_HUMAN | 0.9066 |
| 21 | (+)-catechin | P10275 | ANDR_HUMAN | 0.9896 |
| 22 | (+)-catechin | P00751 | CFAB_HUMAN | 0.9833 |
| 23 | (+)-catechin | P08631 | HCK_HUMAN | 0.9827 |
| 24 | (+)-catechin | P11309 | PIM1_HUMAN | 0.982 |
| 25 | (+)-catechin | P11511 | CP19A_HUMAN | 0.98 |
| 26 | (+)-catechin | P02766 | TTHY_HUMAN | 0.9791 |
| 27 | (+)-catechin | P00734 | THRB_HUMAN | 0.9761 |
| 28 | (+)-catechin | P56817 | BACE1_HUMAN | 0.9732 |
| 29 | (+)-catechin | P03372 | ESR1_HUMAN | 0.9552 |
| 30 | (+)-catechin | P00390 | GSHR_HUMAN | 0.9544 |
| 31 | quercetin | P11309 | PIM1_HUMAN | 0.989 |
| 32 | quercetin | P08631 | HCK_HUMAN | 0.9854 |
| 33 | quercetin | P03372 | ESR1_HUMAN | 0.9615 |
| 34 | quercetin | P06276 | CHLE_HUMAN | 0.9492 |
| 35 | quercetin | P00734 | THRB_HUMAN | 0.9491 |
| 36 | quercetin | P06401 | PRGR_HUMAN | 0.9356 |
| 37 | quercetin | P00533 | EGFR_HUMAN | 0.9333 |
| 38 | quercetin | P02766 | TTHY_HUMAN | 0.9264 |
| 39 | quercetin | P00918 | CAH2_HUMAN | 0.9077 |
| 40 | quercetin | P04062 | GLCM_HUMAN | 0.8908 |
| 41 | beta-sitosterol | P52895 | AK1C2_HUMAN | 0.9987 |
| 42 | beta-sitosterol | P28482 | MK01_HUMAN | 0.9959 |
| 43 | beta-sitosterol | P49137 | MAPK2_HUMAN | 0.9954 |
| 44 | beta-sitosterol | P11309 | PIM1_HUMAN | 0.9921 |
| 45 | beta-sitosterol | P12643 | BMP2_HUMAN | 0.9914 |
| 46 | beta-sitosterol | P00918 | CAH2_HUMAN | 0.9893 |
| 47 | beta-sitosterol | P10828 | THB_HUMAN | 0.9845 |
| 48 | beta-sitosterol | P02768 | ALBU_HUMAN | 0.9838 |
| 49 | beta-sitosterol | P27338 | AOFB_HUMAN | 0.9779 |
| 50 | beta-sitosterol | P08842 | STS_HUMAN | 0.9734 |
| 51 | picralinal | P08254 | MMP3_HUMAN | 0.9695 |
| 52 | picralinal | P23141 | EST1_HUMAN | 0.9578 |
| 53 | picralinal | Q15078 | CD5R1_HUMAN | 0.8947 |
| 54 | picralinal | P53779 | MK10_HUMAN | 0.8889 |
| 55 | picralinal | P06276 | CHLE_HUMAN | 0.8705 |
| 56 | picralinal | P26196 | DDX6_HUMAN | 0.8705 |
| 57 | picralinal | P00747 | PLMN_HUMAN | 0.8672 |
| 58 | picralinal | P45983 | MK08_HUMAN | 0.8601 |
| 59 | picralinal | P55055 | NR1H2_HUMAN | 0.8567 |
| 60 | picralinal | P00918 | CAH2_HUMAN | 0.8265 |
| 61 | physciondiglucoside | P04062 | GLCM_HUMAN | 0.9965 |
| 62 | physciondiglucoside | Q9BW91 | NUDT9_HUMAN | 0.9911 |
| 63 | physciondiglucoside | Q12884 | SEPR_HUMAN | 0.9909 |
| 64 | physciondiglucoside | P17342 | ANPRC_HUMAN | 0.9886 |
| 65 | physciondiglucoside | P12643 | BMP2_HUMAN | 0.9881 |
| 66 | physciondiglucoside | P08631 | HCK_HUMAN | 0.9771 |
| 67 | physciondiglucoside | Q07343 | PDE4B_HUMAN | 0.9762 |
| 68 | physciondiglucoside | P11309 | PIM1_HUMAN | 0.974 |
| 69 | physciondiglucoside | P08254 | MMP3_HUMAN | 0.9713 |
| 70 | physciondiglucoside | P00915 | CAH1_HUMAN | 0.9709 |
| 71 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P07339 | CATD_HUMAN | 0.986 |
| 72 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P53582 | MAP11_HUMAN | 0.9859 |
| 73 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P09211 | GSTP1_HUMAN | 0.9845 |
| 74 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | Q12884 | SEPR_HUMAN | 0.9818 |
| 75 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P09012 | SNRPA_HUMAN | 0.9816 |
| 76 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P25774 | CATS_HUMAN | 0.9809 |
| 77 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P26196 | DDX6_HUMAN | 0.9808 |
| 78 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P04062 | GLCM_HUMAN | 0.9802 |
| 79 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P06276 | CHLE_HUMAN | 0.9775 |
| 80 | torachrysone-8-O-beta-D-(6’-oxayl)-glucoside | P62937 | PPIA_HUMAN | 0.9774 |
| 81 | luteolin | P08631 | HCK_HUMAN | 0.9897 |
| 82 | luteolin | P03372 | ESR1_HUMAN | 0.9684 |
| 83 | luteolin | P06276 | CHLE_HUMAN | 0.9571 |
| 84 | luteolin | P00734 | THRB_HUMAN | 0.9418 |
| 85 | luteolin | Q12884 | SEPR_HUMAN | 0.9349 |
| 86 | luteolin | P02766 | TTHY_HUMAN | 0.9242 |
| 87 | luteolin | P00918 | CAH2_HUMAN | 0.9239 |
| 88 | luteolin | P11309 | PIM1_HUMAN | 0.8977 |
| 89 | luteolin | Q00534 | CDK6_HUMAN | 0.8816 |
| 90 | luteolin | P00533 | EGFR_HUMAN | 0.8601 |
| 91 | emodin | P45983 | MK08_HUMAN | 0.974 |
| 92 | emodin | P02652 | APOA2_HUMAN | 0.9728 |
| 93 | emodin | P08842 | STS_HUMAN | 0.9697 |
| 94 | emodin | P53779 | MK10_HUMAN | 0.9658 |
| 95 | emodin | Q15078 | CD5R1_HUMAN | 0.9655 |
| 96 | emodin | P06276 | CHLE_HUMAN | 0.9651 |
| 97 | emodin | P03372 | ESR1_HUMAN | 0.9642 |
| 98 | emodin | P00918 | CAH2_HUMAN | 0.9526 |
| 99 | emodin | P02768 | ALBU_HUMAN | 0.9517 |
| 100 | emodin | P00915 | CAH1_HUMAN | 0.9411 |
| 101 | chrysophanol | P08254 | MMP3_HUMAN | 0.9971 |
| 102 | chrysophanol | P56817 | BACE1_HUMAN | 0.9635 |
| 103 | chrysophanol | P00751 | CFAB_HUMAN | 0.961 |
| 104 | chrysophanol | P45983 | MK08_HUMAN | 0.9487 |
| 105 | chrysophanol | P06276 | CHLE_HUMAN | 0.9482 |
| 106 | chrysophanol | P53779 | MK10_HUMAN | 0.9449 |
| 107 | chrysophanol | P62937 | PPIA_HUMAN | 0.9429 |
| 108 | chrysophanol | P00734 | THRB_HUMAN | 0.9402 |
| 109 | chrysophanol | Q16539 | MK14_HUMAN | 0.9333 |
| 110 | chrysophanol | P03372 | ESR1_HUMAN | 0.9302 |
| 111 | resveratrol | P08631 | HCK_HUMAN | 0.9869 |
| 112 | resveratrol | P02652 | APOA2_HUMAN | 0.9784 |
| 113 | resveratrol | P03372 | ESR1_HUMAN | 0.9764 |
| 114 | resveratrol | P08842 | STS_HUMAN | 0.9671 |
| 115 | resveratrol | P56817 | BACE1_HUMAN | 0.9609 |
| 116 | resveratrol | P11309 | PIM1_HUMAN | 0.9573 |
| 117 | resveratrol | P00918 | CAH2_HUMAN | 0.9552 |
| 118 | resveratrol | P02768 | ALBU_HUMAN | 0.9518 |
| 119 | resveratrol | P20248 | CCNA2_HUMAN | 0.8837 |
| 120 | resveratrol | P02766 | TTHY_HUMAN | 0.8764 |
| 121 | polydatin | Q9BW91 | NUDT9_HUMAN | 0.991 |
| 122 | polydatin | P54760 | EPHB4_HUMAN | 0.9847 |
| 123 | polydatin | P02652 | APOA2_HUMAN | 0.9802 |
| 124 | polydatin | P08631 | HCK_HUMAN | 0.9783 |
| 125 | polydatin | Q13126 | MTAP_HUMAN | 0.9767 |
| 126 | polydatin | P03372 | ESR1_HUMAN | 0.9762 |
| 127 | polydatin | Q9NP99 | TREM1_HUMAN | 0.9762 |
| 128 | polydatin | P04062 | GLCM_HUMAN | 0.976 |
| 129 | polydatin | P09211 | GSTP1_HUMAN | 0.9739 |
| 130 | polydatin | P42574 | CASP3_HUMAN | 0.9716 |
| 131 | apigenin | P08631 | HCK_HUMAN | 0.9931 |
| 132 | apigenin | P03372 | ESR1_HUMAN | 0.981 |
| 133 | apigenin | P11309 | PIM1_HUMAN | 0.9718 |
| 134 | apigenin | P06276 | CHLE_HUMAN | 0.9705 |
| 135 | apigenin | P04062 | GLCM_HUMAN | 0.9521 |
| 136 | apigenin | Q00534 | CDK6_HUMAN | 0.9 |
| 137 | apigenin | P02766 | TTHY_HUMAN | 0.8903 |
| 138 | apigenin | P00918 | CAH2_HUMAN | 0.8389 |
| 139 | apigenin | P27338 | AOFB_HUMAN | 0.7953 |
| 140 | apigenin | Q92731 | ESR2_HUMAN | 0.773 |
Enrichment analysis and systematic network construction for the putative protein targets
To explore the potential action mechanism of polygonum cuspidatum, GO functional and KEGG pathway enrichment analyses were performed for these 57 targets using OmicShare online tools and DAVID Bioinformatics Resources 6.8, respectively. Based on P < 0.05, the top 5 terms associated with gene counts in biological process were metabolic process, cellular process, single-organism process, response to stimulus and biological regulation. The top 5 terms associated with gene counts in cellular component were cell, cell part, organelle, extracellular region, and membrane. The top 5 terms associated with gene counts in molecular function were binding, catalytic activity, molecular transducer activity, signal transducer activity, and molecular function regulator. These predictive results showed a great diversity in biological functions of polygonum cuspidatum (Figure 2A). Based on P < 0.05, the top 8 KEGG pathways associated with gene counts were pathways in cancer, tuberculosis, mitogen-activated protein kinase 1 (MAPK) signaling pathway, prolactin signaling pathway, progesterone-mediated oocyte-maturation pathway, tumor necrosis factor (TNF) signaling pathway, pathways in hepatitis B and viral carcinogenesis. Of note, we found that the above pathways were closely associated with cancer and gynecological disorders. In fact, previous studies have shown that certain active ingredients in polygonum cuspidatum, such as polydatin, resveratrol and emodin, exhibit potential therapeutic effects against these diseases [4,19-21] (See Table 4 and Figure 2B for details). Further, to provide a systematic overview of the involved signaling pathways and the potential molecular action mechanism of polygonum cuspidatum, we constructed a comprehensive network composed of drug, protein targets and pathways using Cytoscape 3.2.1, based on the results of the above target prediction and pathway enrichment analysis (Figure 2C). The red diamond, orange oval and green rectangle represent 14 active ingredients, protein targets and associated pathways, respectively.
Figure 2.
GO and KEGG enrichment analyses, and systematic network construction. A. GO enrichment analysis map of polygonum cuspidatum potential drug targets by Omicshare tools, containingBiological Process (BP), Cellular Component (CC) and Molecular Function (MF). B. KEGG enrichment analysis map of polygonum cuspidatum potential drug targets by DAVID Bioinformatics Resources 6.8. The size of the dots indicated the number of enriched genes, and the color of the dots represented the degree of significance based on the p-value. C. Drug-target-pathway network construction by Cytoscape 3.2.1. Red diamond: fourteen active ingredients in polygonum cuspidatum; Orange oval: protein targets; Green rectangle: pathways.
Table 4.
The potential pathways targeted by 14 active ingredients in Polygonum cuspidatum
| No. of pathway | Pathway name | Gene count | P-Value |
|---|---|---|---|
| P1 | hsa05200: Pathways in cancer | 8 | 0.014268208 |
| P2 | hsa05152: Tuberculosis | 7 | 0.001037853 |
| P3 | hsa04010: MAPK signaling pathway | 7 | 0.006458276 |
| P4 | hsa04917: Prolactin signaling pathway | 6 | 9.82E-05 |
| P5 | hsa04914: Progesterone-mediated oocyte maturation | 6 | 2.57E-04 |
| P6 | hsa04668: TNF signaling pathway | 6 | 6.42E-04 |
| P7 | hsa05161: Hepatitis B | 6 | 0.002604716 |
| P8 | hsa05203: Viral carcinogenesis | 6 | 0.011172984 |
| P9 | hsa05212: Pancreatic cancer | 5 | 8.65E-04 |
| P10 | hsa05120: Epithelial cell signaling in Helicobacter pylori infection | 5 | 9.70E-04 |
| P11 | hsa05133: Pertussis | 5 | 0.001479731 |
| P12 | hsa04912: GnRH signaling pathway | 5 | 0.003011311 |
| P13 | hsa05145: Toxoplasmosis | 5 | 0.007589309 |
| P14 | hsa04071: Sphingolipid signaling pathway | 5 | 0.008046143 |
| P15 | hsa04722: Neurotrophin signaling pathway | 5 | 0.008046143 |
| P16 | hsa05160: Hepatitis C | 5 | 0.011461535 |
| P17 | hsa04068: FoxO signaling pathway | 5 | 0.01175749 |
| P18 | hsa05164: Influenza A | 5 | 0.027878996 |
| P19 | hsa05169: Epstein-Barr virus infection | 5 | 0.036826594 |
| P20 | hsa05205: Proteoglycans in cancer | 5 | 0.043176273 |
| P21 | hsa04621: NOD-like receptor signaling pathway | 4 | 0.005687485 |
| P22 | hsa05210: Colorectal cancer | 4 | 0.007937757 |
| P23 | hsa05131: Shigellosis | 4 | 0.008664099 |
| P24 | hsa04664: Fc epsilon RI signaling pathway | 4 | 0.010230966 |
| P25 | hsa05132: Salmonella infection | 4 | 0.017504498 |
| P26 | hsa04012: ErbB signaling pathway | 4 | 0.019825453 |
| P27 | hsa04915: Estrogen signaling pathway | 4 | 0.02776564 |
| P28 | hsa05231: Choline metabolism in cancer | 4 | 0.029231619 |
| P29 | hsa04723: Retrograde endocannabinoid signaling | 4 | 0.029231619 |
| P30 | hsa05142: Chagas disease (American trypanosomiasis) | 4 | 0.031506722 |
| P31 | hsa04620: Toll-like receptor signaling pathway | 4 | 0.033074068 |
| P32 | hsa04114: Oocyte meiosis | 4 | 0.035500697 |
| P33 | hsa04930: Type II diabetes mellitus | 3 | 0.040291089 |
| P34 | hsa05150: Staphylococcus aureus infection | 3 | 0.049852736 |
PPI network construction and core gene enrichment analysis
PPI modulation has been attracting increasing attention, and the properties of PPI, such as all osteric sites and hotspots, have been considered by various modern drug-design strategies. We therefore visualized the potential interactions between the 57 protein targets by constructing a PPI network using degree, a key topological parameter, with ‘DC’ scores of ≥ 10 as the cut-off criteria. In total, 50 nodes and 176 edges were generated in this network (Figure 3). The warm color nodes represented a higher degree of core, compared to the cold color nodes within the network. According to the ‘DC’ levels, the 11 core nodes were: Serum albumin (ALB; degree, 34), epidermal growth factor receptor (EGFR; degree, 22), ESR1 (degree, 21), MAPK1 (degree, 19), prothrombin (F2; degree, 17), caspase-3 (CASP3; degree, 17), MAPK8 (degree, 16), MAPK14 (degree, 16), progesterone receptor (degree, 14), androgen receptor (degree, 12) and plasminogen (degree, 12) (Table 5).
Figure 3.
PPI network construction and core targets excavation. The dark nodes represented the core position in this network. The 11 core targets were excavated according to the Degree. PPI, protein-protein interaction.
Table 5.
Hub genes of the PPI network with higher node degrees
| Hub gene | Protein name | Node degree |
|---|---|---|
| ALB | Serum albumin | 34 |
| EGFR | Epidermal growth factor receptor | 22 |
| ESR1 | Estrogen receptor | 21 |
| MAPK1 | Mitogen-activated protein kinase 1 | 19 |
| F2 | Prothrombin | 17 |
| CASP3 | Caspase-3 | 17 |
| MAPK8 | Mitogen-activated protein kinase 8 | 16 |
| MAPK14 | Mitogen-activated protein kinase 14 | 16 |
| PGR | Progesterone receptor | 14 |
| AR | Androgen receptor | 12 |
| PLG | Plasminogen | 12 |
To evaluate the reliability of these core targets, we conducted another core target prediction using a novel formula, which takes the average shortest path length and closeness centrality as the key topological parameters. According to the R-value calculated by this formula, the top 10 core targets of polygonum cuspidatumare presented in Table 6. Of note, 7 core targets, namely EGFR, MAPK1, CASP3, ALB, MAPK14, F2 and ESR1, were overlapping in both topological analyses, demonstrating a good consistency among the different prediction algorithms.
Table 6.
Top 10 targets of polygonum cuspidatum with average shortest path length and closeness centrality
| Protein code | Genes | Description | Average shortest path length | Closeness centrality | R |
|---|---|---|---|---|---|
| P17342 | NPR3 | Atrial natriuretic peptide receptor 3 | 1.000000 | 1.000000 | 0.000000 |
| P02768 | ALB | Serum albumin | 1.260869 | 0.793103 | 0.122449 |
| P04062 | GBA | Lysosomal acid glucosylceramidase, Lysosomal acid GCase | 1.500000 | 0.666667 | 0.234694 |
| Q07343 | PDE4B | cAMP-specific 3’,5’-cyclic phosphodiesterase 4B | 1.500000 | 0.666667 | 0.234694 |
| P00533 | EGFR | Epidermal growth factor receptor | 1.608695 | 0.621622 | 0.285714 |
| P03372 | ESR1 | Estrogen receptor | 1.608695 | 0.621622 | 0.285714 |
| P28482 | MAPK1 | Mitogen-activated protein kinase 1, MAP kinase 1, MAPK 1 | 1.695652 | 0.589744 | 0.326531 |
| P42574 | CASP3 | Caspase-3, CASP-3 | 1.695652 | 0.589744 | 0.326531 |
| P00734 | F2 | Prothrombin | 1.695652 | 0.589744 | 0.326531 |
| Q16539 | MAPK14 | Mitogen-activated protein kinase 14, MAP kinase 14, MAPK 14 | 1.739130 | 0.575000 | 0.346939 |
Identification of the key ingredients in polygonum cuspidatum by UPLC-MS/MS and prediction of their potential drug targets
To validate the 14 key ingredients previously selected based on the ADME-related characteristics and literature search, we next set out to identify the ingredients from the crude extract of polygonum cuspidatum using the UPLC-MS/MS method. Seven main ingredients of polygonum cuspidatum were identified, including resveratrol (tR: 11.89 min), emodin (tR: 23.70 min), apigenin (tR: 17.52 min), rhein (tR: 19.54 min), catechin (tR: 10.37 min), quercetin (tR: 17.08 min) and polydatin (tR: 11.73 min) (Figure 4A-D; see Table 7 for details).
Figure 4.
Seven main ingredients obtained from the crude extract of polygonum cuspidatum by using the UPLC-MS/MS method. A. Ultraviolet absorption wavelength map. B. Total Ion Chromatography (TICs) in positive ion modes. C. Total Ion Chromatography (TICs) in negative ion modes. D. Identification maps of the 7 main ingredients by secondary mass spectrometry. E. Venn diagram (6 core targets from 7 main ingredients and topological analysis). F. KEGG enrichment analysis map of 6 core targets by DAVID Bioinformatics Resources 6.8. The size of the dots indicated the number of enriched genes, and the color of the dots represented the degree of significance based on the p-value.
Table 7.
Identification of compounds in Polygonum cuspidatum by UPLC-MS data
| No. | tR (min) | Molecular formula | Selected ion | Theoretical | Experimental | MS/MS fragmentions | Compounds |
|---|---|---|---|---|---|---|---|
| 1 | 11.89 | C14H12O3 | [M+H]+ | 228.24 | 229.24 | 211.07, 183.11, 165.01, 134.89, 118.94, 106.80, 90.85 | Resveratrol |
| 2 | 23.70 | C15H10O5 | [M-H]- | 270.24 | 269.32 | 241.24, 225.30, 181.72 | Emodin |
| 3 | 17.52 | C15H10O5 | [M-H]- | 270.24 | 269.51 | 159.10, 132.86, 106.99, 63.14 | Apigenin |
| 4 | 19.54 | C15H8O6 | [M-H]- | 284.22 | 283.36 | 240.29 | Rhein |
| 5 | 10.37 | C15H14O6 | [M+H]+ | 290.27 | 291.04 | 138.57, 123.01 | Catechin |
| 6 | 17.08 | C15H10O7 | [M-H]- | 302.24 | 301.05 | 283.64, 152.06, 124.99, 106.70, 82.46 | Quercetin |
| 7 | 11.73 | C20H22O8 | [M-H]- | 390.39 | 389.53 | 227.36 | Polydatin |
The potential drug targets for each of these 7 ingredients were then predicted using PharmMapper, and a total of 38 potential protein targets for these 7 ingredients were finally obtained. Interestingly, among these targets, 6 overlapped with the 7 core targets excavated from the previous topological analyses, including EGFR, CASP3, ALB, MAPK14, F2 and ESR1, demonstrating that the network pharmacology approach is a reliable and powerful tool for the study of the CHM mechanism (Figure 4E).
Subsequent pathway enrichment analysis revealed that these 6 core genes were mainly enriched in proteoglycans in cancer and the MAPK signaling pathway (Figure 4F). Meanwhile, other pathways, such as the prolactin, estrogen and TNF signaling pathways, as well as pathways in cancer, were also enriched in this analysis, which was consistent with the results of the aforementioned pathway enrichment analysis for the 57 putative targets. These results suggested that these core genes are closely correlated with cancer and gynecological diseases, which was also in line with previous reports [22,23].
Polygonum cuspidatum decreased the cell viability and induced the apoptosis of MDA-MB-231 and SKOV-3 cells
To validate the potential therapeutic effects of polygonum cuspidatum on cancer and gynecological disorders, we next carried out a series of cell functional assays, using human breast cancer MDA-MB-231 and ovarian cancer SKOV-3 cells as the experimental models of these diseases. The results of the CCK-8 assay revealed that the viability of both cells was significantly inhibited by polygonum cuspidatum in a dose and time dependent manner (Figure 5A, 5B). After treating these cells for 24 h, IC50 analysis indicated that polygonum cuspidatum exerted its 50% inhibitory effect on MDA-MB-231 cells at 31.18 ± 1.95 μg/ml and SKOV-3 cells at 28.12 ± 1.07 μg/ml, respectively. Based on the above IC50 values, we next evaluated the apoptotic ratio of these cells upon polygonum cuspidatum treatment. The flow cytometry results showed that the apoptotic population stained with Annexin V-FITC was dramatically increased following the treatment (Figure 5C). Furthermore, the immunoblotting results revealed that some of the key factors associated with MAPK and TNF signaling pathways in these cells, such as p-ERK (T202/Y204), NF-κB (p65) and IKK α/β, were significantly decreased in a dose dependent manner upon polygonum cuspidatum treatment, which was consistent with our previous KEGG enrichment analysis (Figure 5D).
Figure 5.
Polygonum cuspidatum decreased the viability and induced apoptosis of MDA-MB-231 and SKOV-3 cells. A, B. CCK-8 assay was performed to measure cell proliferation rate at 24 and 48 h after polygonum cuspidatum treatment at different dosages (0, 25, 50, 100, 200 and 400 μg/ml). C. Induction of apoptosis of MDA-MB-231 and SKOV-3 cells after polygonum cuspidatum treatment. MDA-MB-231 cells were treated with polygonum cuspidatum at different concentrations (0 and 31 μg/ml) for 24 h, and SKOV-3 cells were also treated with polygonum cuspidatum at different concentrations (0 and 28 μg/ml) for 24 h, when apoptotic events were assessed by flow cytometry. Statistical analysis of the percentage of apoptosis in MDA-MB-231 and SKOV-3 cells **P < 0.01 based on Student’s t-test. D. MDA-MB-231 and SKOV-3 cells were treated with polygonum cuspidatum at different concentrations (0, 31, 62 μg/ml or 0, 28, 56 μg/ml) for 24 h. After proteins were extracted, the protein levels of p-ERK (T202/Y204), ERK, NF-κB, IKK-α and IKK-β were analyzed by western blot.
Next, the levels of the aforementioned 6 core targets were determined upon polygonum cuspidatum treatment. The quantitative real time polymerase chain reaction (qRT-PCR) results revealed that the mRNA levels of EGFR, CASP3 and F2 were significantly downregulated by polygonum cuspidatum in SKOV-3 cells (P < 0.05; Figure 6A). Consistently, the protein level of EGFR was down-regulated by polygonum cuspidatum in a dose-dependent manner. Meanwhile, the total protein level of CASP3 was also slightly decreased, while the cleaved CASP3, a well-known apoptotic maker, accumulated significantly upon polygonum cuspidatum treatment (Figure 6B).
Figure 6.
The mRNA and protein expression levels of six core targets. A. The mRNA expression levels of 6 core targets were analyzed by qRT-PCR in human ovarian cancer SKOV-3 cells. **P < 0.01 based on Student’s t-test. qRT-PCR, quantitative real time polymerase chain reaction. B. SKOV-3 cells were treated with polygonum cuspidatum at different concentrations (0, 28, 56 μg/ml) for 24 h. After proteins were extracted, the protein levels of EGFR and CASP3 were analyzed by western blot.
Systemsdock analysis of key ingredients and core targets
To further explore the action mechanism of polygonum cuspidatum at the protein level, we performed molecular docking analysis for its key ingredients and the corresponding core targets using SystemsDock, a unique web server that applies docking simulation for the comprehensive characterization of ligand selectivity and interpretation of ligand action on a complex molecular network [24,25]. The docking efficiency between the UPLC-MS/MS derived 7 key ingredients (quercetin, polydatin, resveratrol, emodin, apigenin, rhein and catechin) and the 6 core targets (EGFR, CASP3, ALB, MAPK14, F2 and ESR1) were predicted. The results showed that the overall binding affinity of polydatin with the 6 core targets (scores: 6.78, 6.58, 6.74, 6.63, 6.50 and 6.19, respectively) was greater than those of the other 6 ingredients (Figure 7A-G). Therefore, this result suggested that polydatin possesses a potential pharmacological activity.
Figure 7.
Molecular docking analysis by SystemsDock. Scores of molecular docking between 6 core targets and (A) quercetin, (B) polydatin, (C) resveratrol, (D) emodin, (E) apigenin, (F) rhein and (G) catechin, respectively. Black column: native ligand; Red column: seven main ingredients.
Next, we further explored the druggability of polydatin using Autodock. Since our previous prediction showed that CASP3 was both a direct target of polydatin and one of the core targets of polygonum cuspidatum, CASP3 was selected as an example for the next docking analysis. The results showed that the physical and chemical properties of polydatin in the protein binding pocket were stable due to its rich hydroxyl groups with a good hydrophilic property. Meanwhile, the surrounding amino acids interacting with polydatin were all polar amino acids, such as Ser, Thr, Arg, Glu and Gln, which could form a strong hydrogen bond between polydatin and the protein binding pocket of CASP3 (Figure 8). Thus, polydatin may play a pivotal role in manifesting the therapeutic effects of polygonum cuspidatum via binding with CASP3. However, further experimental verification is warranted.
Figure 8.
Binding modes from Autodock. Hydrophilic-hydrophobic interaction between (A) Polydatin and CASP3 in protein binding pocket, and (B) Polydatin and relative amino acids. CASP3, caspase-3.
Discussion
Nowadays, a myriad of in silico target predicting methods have been used in the field of drug development [26]. In the present study, the PharmMapper database, which is based on 7,000 pharmacophore models, was utilized to search for the potential targets for the key ingredients of polygonum cuspidatum. The subsequent GO functional enrichment analysis of the 57 potential targets revealed a great variety of biological functions of polygonum cuspidatum. Furthermore, the KEGG pathway enrichment results showed that these targets were mainly involved in carcinogenesis and gynecological diseases. Indeed, accumulating evidence have shown that polygonum cuspidatum and its active ingredients exhibit anti-cancer effects either through direct inhibition of cancer cell growth or through the modulation of the host immune-system [27-31]. In addition, recent pharmacological studies showed that polygonum cuspidatum extractor its major ingredients possessed anti-virus and anti-inflammatory activities, which may, at least partially, reflect the therapeutic effects of polygonum cuspidatum on those cancerous and gynecological diseases attributed to viral infection or inflammation [32-35]. Of note, different active ingredients of polygonum cuspidatum, such as resveratrol and polydatin, may exert their therapeutic functions via distinct mechanisms. For instance, polydatin could inhibit cell growth and induce the apoptosis of HCC cells by blocking the STAT3 signaling pathway, while resveratrol inhibiting the epithelial mesenchymal transition through inactivation of the PI-3K/Akt/NF-κB and hedgehog signaling pathways [36-39]. Therefore, the present action mechanism study guided by network pharmacology may help provide a deeper insight into the “multi-component, multi-target and multi-function” characteristics of polygonum cuspidatum in cancerous and gynecological diseases.
To untangle the complicated interactions among the potential targets of the key ingredients in polygonum cuspidatum, a PPI network was constructed based on a key topological parameter, degree, using STRING and Cytoscape 3.2.1. Eleven core targets were found to have higher degrees of connectivity within this network. Besides the degree, other topological parameters, such as shortest path length and closeness centrality, could also be used for assessing the importance of key nodes (core targets) within a network. The nodes with a small average shortest path length and a large closeness centrality (R-value) were usually considered as the important targets in the network [40]. We therefore developed a novel formula, in which the “R value” is determined by the average shortest path length and closeness centrality, to validate the reliability of the above 11 predicted core targets. Notably, among the top 10 core targets (ranked based on R-values) excavated using this new formula, 7 overlapped with the former 11 core targets, showing a good consistency among the different prediction algorithms.
Noteworthy, 7 ingredients of polygonum cuspidatum were validated by UPLC-MS/MS, among the 14 putative ones derived from the network pharmacological prediction. Of these 7 ingredients, only 3 met the medicinal criteria for the selection of active ingredients (OB ≥ 30% and DL ≥ 0.18), indicating the importance of literature mining to the selection of active ingredients by network pharmacology. Interestingly, 6 potential targets derived from these 7 ingredients overlapped with the 7 core targets excavated from the previous PPI network analyses, underscoring the pivotal role of these 6 targets in the action mechanism of polygonum cuspidatum. Our further molecular docking analysis between these 6 targets and 7 ingredients showed that polydatin possesses a better pharmacological activity, as compared to the other 6 ingredients, implying a leading role of polydatin in delivering the therapeutic effects of polygonum cuspidatum. Indeed, among the 6 aforementioned potential targets of polygonum cuspidatum, EGFR and CASP3 have previously been identified as the putative targets of polydatin using PharmMapper and ChemMapper algorithms (7). In the present study, we confirmed that EGFR and CASP3 were down-regulated at both mRNA and protein levels by qRT-PCR and western blot in polygonum cuspidatum-treated cells, accompany with the accumulation of the cleaved CASP3, which was in line with the apoptotic phenotype induced by polydatin in this and other studies [36,41]. These results further suggested that polydatin could exert its inhibitory effects on these two targets at both mRNA and protein levels. Nevertheless, polydatin exhibited its antiproliferation effect on different cancerous cells via distinct mechanisms. For instance, polydatin could arrest the leukemia cells in the S-phase by inhibiting the expression of cyclins D1 and B1, while inhibiting the proliferation of breast cancer cells by inactivating Creb [42,43]. Therefore, the detailed action mechanism of polydatin underlying its therapeutic effect on specific cell types warrants further investigation.
Acknowledgements
This study was funded by the National Natural Science Foundation of China (grant no. 81572416), the National Key Technologies R & D Program of China (grant no. 2016YFC1303200) and the Tianjin Medical University Cancer Institute & Hospital Cancer Translational Medicine Seed Funds (grant no. 1701-1).
Disclosure of conflict of interest
None.
Abbreviations
- CHM
Chinese herb medicine
- PPI
protein-protein interaction
- OB
oral bioavailability
- DL
drug-likeness
- EGFR
epidermal growth factor receptor
- CASP3
caspase-3
- TICs
Total Ion Chromatography
References
- 1.Gao H, Wang Z, Li Y, Qian Z. Overview of the quality standard research of traditional Chinese. Front Med. 2011;5:195–202. doi: 10.1007/s11684-011-0134-x. [DOI] [PubMed] [Google Scholar]
- 2.Tang WM, Chan E, Kwok CY, Lee YK, Wu JH, Wan CW, Chan RY, Yu PH, Chan SW. A review of the anticancer and immunomodulatory effects of Lycium barbarum fruit. Inflammopharmacology. 2011;20:307–14. doi: 10.1007/s10787-011-0107-3. [DOI] [PubMed] [Google Scholar]
- 3.Chan SW. Panax ginseng, rhodiolarosea and schisandra chinensis. Int J Food Sci Nutr. 2012;2012:75–81. doi: 10.3109/09637486.2011.627840. [DOI] [PubMed] [Google Scholar]
- 4.Peng W, Qin R, Li X, Zhou H. Botany, phytochemistry, pharmacology, and potential application of polygonum cuspidatum Sieb.et Zucc.: a review. J Ethnopharmacol. 2013;148:729–45. doi: 10.1016/j.jep.2013.05.007. [DOI] [PubMed] [Google Scholar]
- 5.Matsuda H, Shimoda H, Morikawa T, Yoshikawa M. Phytoestrogens from the roots of polygonum cuspidatum (polygonaceae): structure-requirement of hydroxyanthraquinones for estrogenic activity. Bioorg Med Chem Lett. 2001;11:1839–42. doi: 10.1016/s0960-894x(01)00318-3. [DOI] [PubMed] [Google Scholar]
- 6.Liu H, Zhan S, Zhang Y, Ma Y, Chen L, Chen L, Dong H, Ma M, Zhang Z. Molecular network-based analysis of the mechanism of liver injury induced by volatile oils from Artemisiaeargyi folium. BMC Complement Alteru Med. 2017;17:491. doi: 10.1186/s12906-017-1997-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Pan BY, Ren YY, Liu LR. Uncovering the action mechanism of polydatin via network pharmacological target prediction. RSC Adv. 2018;8:18851. doi: 10.1039/c8ra03124j. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hong M, Li S, Tan HY, Cheung F, Wang N, Huang J, Feng Y. A network-based pharmacology study of the herb-induced liver injury potential of traditional hepatoprotective Chinese herbal medicines. Molecules. 2017;22:632–46. doi: 10.3390/molecules22040632. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Lyu M, Yan CL, Liu HX, Wang TY, Shi XH, Liu JP, Orgah J, Fan GW, Han JH, Wang XY, Zhu Y. Network pharmacology exploration reveals endothelial inflammation as a common mechanism for stroke and coronary artery disease treatment of Danhong injection. Sci Rep. 2017;7:15427–45. doi: 10.1038/s41598-017-14692-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zhang A, Fang H, Wang Y, Yan G, Sun H, Zhou X, Wang Y, Liu L, Wang X. Discovery and verification of the potential targets from bioactive molecules by network pharmacology-based target prediction combined with high-throughput metabolomics. RSC Adv. 2017;7:51069–78. [Google Scholar]
- 11.Hong M, Zhang Y, Li S, Tan HY, Wang N, Mu S, Hao X, Feng Y. A network pharmacology-based study on the hepatoprotective effect of fructus schisandrae. Molecules. 2016;22:1617–28. doi: 10.3390/molecules22101617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pei T, Zheng C, Huang C, Chen X, Guo Z, Fu Y, Liu J, Wang Y. Systematic understanding the mechanisms of vitiligo pathogenesis and its treatment by qubaibabuqi formula. J Ethnopharmacol. 2016;190:272–87. doi: 10.1016/j.jep.2016.06.001. [DOI] [PubMed] [Google Scholar]
- 13.Zhang W, Tao Q, Guo Z, Fu Y, Chen X, Shar PA, Shahen M, Zhu J, Xue J, Bai Y, Wu Z, Wang Z, Xiao W, Wang Y. Systems pharmacology dissection of the integrated treatment for cardiovascular and gastrointestinal disorders by traditional Chinese medicine. Sci Rep. 2016;6:32400. doi: 10.1038/srep32400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li B, Weng Q, Dong C, Zhang Z, Li R, Liu J, Jiang A, Li Q, Jia C, Wu W, Liu H. A key gene, PLIN1, can affect porcine intramuscular fat content based on transcriptome analysis. Genes. 2003;9:194. doi: 10.3390/genes9040194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.de Jong H, Geiselmann J, Hernandez C, Page M. Genetic network analyzer: qualitative simulation of genetic regulatory networks. Bioinformatics. 2003;19:336–44. doi: 10.1093/bioinformatics/btf851. [DOI] [PubMed] [Google Scholar]
- 16.Ru J, Li P, Wang J, Zhou W, Li B, Huang C, Li P, Guo Z, Tao W, Yang Y, Xu X, Li Y, Wang Y, Yang L. Tcmsp: a database of systems pharmacology for drug discovery from herb medicines. J Cheminf. 2014;6:13. doi: 10.1186/1758-2946-6-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Shi S, Pan M, Wang W. Research progress on chemical in roots and rhizomes of polygonum cuspidatum and their pharmacological activities. Drug Evaluation Research (Chinese Journal) 2016;39:317–21. [Google Scholar]
- 18.Kong X, Zhou L. Study progress of giant knotweed rhizome. Guiding Journal of Traditional Chinese Medicine and Pharmacy (Chinese Journal) 2009;15:107–10. [Google Scholar]
- 19.Chen S, Tao J, Zhong F, Jiao Y, Xu J, Shen Q, Wang H, Fan S, Zhang Y. Polydatin down-regulates the phosphorylation level of Creb and induces apoptosis in human breast cancer cell. PLoS One. 2017;12:e0176501. doi: 10.1371/journal.pone.0176501. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Li H, Shi B, Li Y, Yin F. Polydatin inhibits cell proliferation and induces apoptosis in laryngeal cancer and HeLa cells via suppression of the PDGF/AKT signaling pathway. Biochem Mol Toxicol. 2017;31:e21900. doi: 10.1002/jbt.21900. [DOI] [PubMed] [Google Scholar]
- 21.Díaz-Chávez J, Fonseca-Sánchez MA, Arechaga-Ocampo E, Flores-Pérez A, Palacios-Rodríguez Y, Domínguez-Gómez G, Marchat LA, Fuentes-Mera L, Mendoza-Hernández G, Gariglio P, López-Camarillo C. Proteomic profiling reveals that resveratrol inhibits HSP27 expression and sensitizes breast cancer cells to doxorubicin therapy. PLoS One. 2013;8:e64378. doi: 10.1371/journal.pone.0064378. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Park SH, Seong MA, Lee HY. p38 MAPK-induced MDM2 degradation confers paclitaxel resistance through p53-mediated regulation of EGFR in human lung cancer cells. Oncotarget. 2016;7:8184–99. doi: 10.18632/oncotarget.6945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.O’Leary KA, Jallow F, Rugowski DE, Sullivan R, Sinkevicius KW, Greene GL, Schuler LA. Prolactin activates ERα in the absence of ligand in female mammary development and carcinogenesis in vivo. Endocrinology. 2013;154:4483–92. doi: 10.1210/en.2013-1533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hsin KY, Matsuoka Y, Asai Y, Kamiyoshi K, Watanabe T, Kawaoka Y, Kitano H. Systemsdock a web server for network pharmacology-based prediction and analysis. Nucleic Acids Res. 2016;44:W507–13. doi: 10.1093/nar/gkw335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hsin KY, Ghosh S, Kitano H. Combining machine learning systems and multiple docking simulation packages to improve docking prediction reliability for network pharmacology. PLoS One. 2013;8:e83922. doi: 10.1371/journal.pone.0083922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cereto-Massagué A, Ojeda MJ, Valls C, Mulero M, Pujadas G, Garcia-Vallve S. Tools for in silico target fishing. Methods. 2015;71:98–103. doi: 10.1016/j.ymeth.2014.09.006. [DOI] [PubMed] [Google Scholar]
- 27.Wu X, Li Q, Feng Y, Ji Q. Antitumor research of the active ingredients from traditional Chinese medical plant polygonum cuspidatum. Evid Based Complement Alternat Med. 2018;2018:2313021. doi: 10.1155/2018/2313021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Li W, Zhang Q, Chen K, Sima Z, Liu J, Yu Q, Liu J. 2-Ethoxystypandrone, a novel small-molecule STAT3 signaling inhibitor from polygonum cuspidatum, inhibits cell growth and induces apoptosis of HCC cells and HCC Cancer stem cells. BMC Complement Altern Med. 2019;19:38. doi: 10.1186/s12906-019-2440-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wang YL, Horng CT, Hsieh MT, Chen HC, Huang YS, Yang JS, Wang GK, Chiang JH, Chen HH, Lu CC, Chen FA. Autophagy and apoptotic machinery caused by polygonum cuspidatum extract in cisplatin resistant human oral cancer CAR cells. Oncol Rep. 2019;41:2549–57. doi: 10.3892/or.2019.6985. [DOI] [PubMed] [Google Scholar]
- 30.Chueh FS, Lin JJ, Lin JH, Weng SW, Huang YP, Chung JG. Crude extract of polygonum cuspidatum stimulates immune responses in normal mice by increasing the percentage of Mac-3-positive cells and enhancing macrophage phagocytic activity and natural killer cell cytotoxicity. Mol Med Rep. 2015;11:127–32. doi: 10.3892/mmr.2014.2739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Chueh FS, Lin JJ, Lin JP, Yu FS, Lin JH, Ma YS, Huang YP, Lien JC, Chung JG. Crude extract of Polygonum cuspidatum promotes immune responses in leukemic mice through enhancing phagocytosis of macrophage and natural killer cell activities in vivo. In Vivo. 2015;29:255–61. [PubMed] [Google Scholar]
- 32.Chang JS, Liu HW, Wang KC, Chen MC, Chiang LC, Hua YC, Lin CC. Ethanol extract of polygonum cuspidatum inhibits hepatitis B virus in a stable HBV-producing cell line. Antivir Res. 2005;66:29–34. doi: 10.1016/j.antiviral.2004.12.006. [DOI] [PubMed] [Google Scholar]
- 33.Lin HW, Sun MX, Wang YH, Yang LM, Yang YR, Huang N, Xuan LJ, Xu YM, Bai DL, Zheng YT, Xiao K. Anti-HIV activities of the compounds isolated from polygonum cuspidatum and polygonum multiflorum. Planta Med. 2010;76:889–92. doi: 10.1055/s-0029-1240796. [DOI] [PubMed] [Google Scholar]
- 34.Wang QW, Su Y, Sheng JT, Gu LM, Zhao Y, Chen XX, Chen C, Li WZ, Li KS, Dai JP. Anti-influenza a virus activity of rhein through regulating oxidative stress, TLR4, Akt, MAPK, and NF-κB signal pathways. PLoS One. 2018;13:e0191793. doi: 10.1371/journal.pone.0191793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Han JH, Koh W, Lee HJ, Lee HJ, Lee EO, Lee SJ, Khil JH, Kim JT, Jeong SJ, Kim SH. Analgesic and anti-inflammatory effects of ethyl acetate fraction of polygonum cuspidatum in experimental animals. Immunopharm Immunot. 2012;34:191–5. doi: 10.3109/08923973.2011.590499. [DOI] [PubMed] [Google Scholar]
- 36.Jiao Y, Wu Y, Du D. Polydatin inhibits cell proliferation, invasion and migration, and induces cell apoptosis in hepatocellular carcinoma. Braz J Med Biol Res. 2018;51:e6867. doi: 10.1590/1414-431X20176867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Li W, Ma J, Ma Q, Li B, Han L, Liu J, Xu Q, Duan W, Yu S, Wang F, Wu E. Resveratrol inhibits the epithelial-mesenchymal transition of pancreatic cancer cells via suppression of the PI-3K/Akt/NF-κB pathway. Curr Med Chem. 2013;20:4185–94. doi: 10.2174/09298673113209990251. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wang H, Zhang H, Tang L, Chen H, Wu C, Zhao M, Yang Y, Chen X, Liu G. Resveratrol inhibits TGF-β1-induced epithelial-to-mesenchymal transition and suppresses lung cancer invasion and metastasis. Toxicology. 2013;303:139–46. doi: 10.1016/j.tox.2012.09.017. [DOI] [PubMed] [Google Scholar]
- 39.Li J, Chong T, Wang Z, Chen H, Li H, Cao J, Zhang P, Li H. A novel anti-cancer effect of resveratrol: reversal of epithelial-mesenchymal transition in prostate cancer cells. Mol Med Rep. 2014;10:1717–24. doi: 10.3892/mmr.2014.2417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Anitha P, Anbarasu A, Ramaiah S. Gene network analysis reveals the association of important functional partners involved in antibiotic resistance: a report on an important pathogenic bacterium staphylococcus aureus. Gene. 2016;575:253–63. doi: 10.1016/j.gene.2015.08.068. [DOI] [PubMed] [Google Scholar]
- 41.Liu H, Zhao S, Zhang Y, Wu J, Peng H, Fan J, Liao J. Reactive oxygen species-mediated endoplasmic reticulum stress and mitochondrial dysfunction contribute to polydatin-induced apoptosis in human nasopharyngeal carcinoma CNE cells. J Cell Biochem. 2011;112:3695–703. doi: 10.1002/jcb.23303. [DOI] [PubMed] [Google Scholar]
- 42.Cao WJ, Wu K, Wang C, Wan DM. Polydatin-induced cell apoptosis and cell cycle arrest are potentiated by Janus kinase 2 inhibition in leukemia cells. Mol Med Rep. 2016;13:3297–302. doi: 10.3892/mmr.2016.4909. [DOI] [PubMed] [Google Scholar]
- 43.Chen S, Tao J, Zhong F, Jiao Y, Xu J, Shen Q, Wang H, Fan S, Zhang Y. Polydatin down-regulates the phosphorylation level of Creb and induces apoptosis in human breast cancer cell. PLoS One. 2017;12:e0176501. doi: 10.1371/journal.pone.0176501. [DOI] [PMC free article] [PubMed] [Google Scholar]