Network Pharmacology to Uncover the Molecular Mechanisms of Action of LeiGongTeng for the Treatment of Nasopharyngeal Carcinoma

Abstract

Background

Nasopharyngeal carcinoma (NPC) is a common head and neck cancer epidemic in southern China and southeast Asia. LeiGongTeng has been widely used for the treatment of cancers. The purpose of this study was to determine the pharmacological mechanism of action of LeiGongTeng in the treatment of NPC using a network pharmacological approach.

Material/Methods

The traditional Chinese medicine systems pharmacology (TCMSP) database was used to identify active ingredients and associated target proteins for LeiGongTeng. Cytoscape was utilized to create a drug-disease network and topology analysis was conducted to analyze the degree of each ingredient. The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) online tool was applied for the construction and analysis of the protein-protein interaction (PPI) network, while Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment and Gene Ontology (GO) functional analyses were utilized to determine drug-disease common genes.

Results

22 active ingredients including kaempferol, nobiletin, and beta-sitosterol, and 30 drug-disease common genes including VEGFA, CASP3, ESR1, and RELA were identified. GO analysis indicated that 94 biological processes, including RNA polymerase II, apoptotic process, response to drug, cell adhesion, and response to hypoxia, were found to be associated with NPC. The KEGG enrichment analysis showed that 58 pathways, including the PI3K-Akt signaling pathway, microRNAs in cancer, tumor necrosis factor (TNF) signaling pathway and pathways in cancer were found to be associated with NPC.

Conclusions

LeiGongTeng exerts its therapeutic effect through various biological processes and signaling pathways since it acts on several target genes. Systematic pharmacology can be used to predict the underlying function of LeiGongTeng and its mechanism of action in NPC.

Background

Nasopharyngeal carcinoma (NPC), a type of head and neck cancer, is an epithelial carcinoma arising from the nasopharyngeal mucosal lining [1]. It has a characteristic specific geographic distribution and racial prevalence [2]. The incidence of NPC is 50 cases per 100 000 people per year in southern China and southeast Asia [3]. Distant metastasis has resulted in the failure of treatment for these patients [4]. Due to its anatomical location, radiotherapy and chemotherapy have been considered standard treatment for patients with NPC [5]. However, side effects, such as radiation-related skin reaction, vomiting, and leukopenia may be particularly concerning for NPC patients [6]. Therefore, more effective and safe strategies for the treatment of NPC are still needed.

In Asia, traditional Chinese medicine (TCM) has been used frequently in cancer treatment [7]. TCM has the advantage of producing a reliable therapeutic efficacy, while inducing fewer adverse effects, and has drawn much attention in western countries during recent years [8]. LeiGongTeng is also known as Tripterygium wilfordii or Tripterygii Radix. Anti-cancer activity of Tripterygii Radix against many kinds of cancers has been illustrated [9–11]. Triptolide is a bio-active component isolated from Tripterygii Radix [12], previous studies reported that triptolide induce Epstein-Barr virus (EBV) nuclear antigen 1 (EBNA1) degradation and stimulate NPC cells apoptosis via mitochondria apoptotic pathway [13], triptolide in combined with cisplatin (DDP) showed a synergistic effect against DDP-resistant in NPC cells [14], moreover, triptolide in combination with ionizing radiation exhibits synergistic effects of anti-cancer and anti-angiogenesis in NPC cells [15]. The complex active compounds of TCM are difficult to clarify, while previous and potential pharmacological mechanisms of Tripterygii Radix in NPC have not been fully elucidated. Network pharmacology is an advanced method based on chemoinformatics, bioinformatics, network biology, and pharmacology [16]. Using a network-based approach, the active ingredients, fundamental molecular mechanisms, and pathways of TCM used for treatment can be systematically identified. In the study by He et al., 33 constituents of the Compound Kushen Injection (CKI) were found to be associated with anti-cancer activity, while 113 targeted proteins, 129 biological processes and 93 related pathways were identified [17]. The report by Yang et al. showed that 146 related proteins of 82 bioactive compounds in Wei Pi Xiao (WPX) decoction may explain the mechanism of activity involved in gastric precancerous lesion treatment, and 21 signaling pathways and 26 key biological processes were identified [18].

The purpose of this study was to analyze molecular mechanisms involved in the treatment effect and the active ingredients of Tripterygii Radix using the network pharmacology approach. The targets and active ingredients of Tripterygii Radix were identified and used to determine those that were common with Tripterygii Radix for the treatment of NPC. Then, construction of the protein-protein interaction (PPI) network, as well as Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) term enrichment analyses were conducted to identify related biological processes and signaling pathways.

Material and Methods

Identification of active compounds

Major compounds of Tripterygii Radix were derived using a database of Chinese herbal medicines that contains information on herbal entries, drug-disease networks and drug-target networks, and the traditional Chinese medicine systems pharmacology (TCMSP) database and analysis platform (http://lsp.nwu.edu.cn/tcmsp.php) [19]. The screening cutoff standards used were drug-likeness (DL) of ≥0.18 and oral bioavailability (OB) of ≥30%, and the compounds that satisfied these criteria were regarded as candidate compounds. The genes corresponding to the targets were obtained from the UniProt database.

Identification of disease target genes

Information on NPC-associated target genes was obtained from the GeneCards database (http://www.genecards.org/) and the Online Mendelian Inheritance in Man database (OMIM, http://www.ncbi.nlm.nih.gov/omim). The GeneCards database is a website that integrates genetic, proteomic, transcriptomic, and genomic information [20]. OMIM was created by McKusick in the early 1960s; it contains an authoritative and comprehensive compendium of human genetic phenotypes and genes [21]. The VennDiagram package of R software was utilized to obtain common target genes for drugs and diseases.

Construction of the “drug-disease” network

In order to comprehensively explore the pharmacological mechanisms involved, the Cytoscape (http://www.cytoscape.org/) visualization software 3.6.1 was used to create the drug-disease network [22]. Targets and compounds were input into Cytoscape and the drug-disease interaction network was established. Furthermore, we used the plug-in, NetworkAnalysizer, to calculate the degree of each compound. Degree is a critical parameter of the topology structure that is used to assess the importance of a compound.

Construction of the PPI network

The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING, https://string-db.org/) is an online database of known and predicted protein-protein interactions [23]. The interactions included indirect (functional) and direct (physical) interactions. The network nodes are proteins and the edges show their associations. To further explore protein interactions systematically, we input common target genes for drugs into STRING to obtain relevant information on protein interactions. Then, this network was exported, and we carried out further statistical analysis of the protein interactions using Cytoscape 3.6.1 software. We used the plugin, Cytohubba, on hub genes from the PPI network in Cytoscape with a degree of >16.

GO analysis and KEGG pathway enrichment analysis

The GO Consortium database provides and limits the functions of genes. Biological processes related to various biological phenomena can be effectively identified through this method [24]. KEGG is a database developed by the University of Tokyo and Kyoto University, Japan, which is applied to screen functional and metabolic pathways [25]. The database for annotation, visualization, and integrated discovery (DAVID) was used to conduct the KEGG enrichment pathway and GO analyses [26]. DAVID is an online database, which can be used for enrichment analysis to show highly associated GO terms and KEGG pathways. We used the ClusterProfiler package of R software to visualize these results.

Results

Screening of active ingredients and target genes

The targets and active ingredients of Tripterygii Radix were predicted using the TCMSP database. We identified 144 compounds of Tripterygium wilfordii in total. Using the cutoff criteria of DL ≥0.18 and OB ≥30%, 51 active ingredients were identified (Table 1). Furthermore, we entered the target name and chose “human” as the species in the UniProt database. In total, 62 genes corresponding the potential targets of these 51 ingredients were identified.

Table 1.

The active ingredients of Tripterygii Radix.

Mol ID	Molecule name	OB (%)	DL	Caco-2	Molecular formula
MOL000211	Mairin	55.38	0.78	0.73	C30H48O3
MOL000296	Hederagenin	36.91	0.75	1.32	C30H48O4
MOL000358	Beta-sitosterol	36.91	0.75	1.32	C29H50O
MOL000422	Kaempferol	41.88	0.24	0.26	C15H10O6
MOL000449	Stigmasterol	43.83	0.76	1.44	C29H48O
MOL002058	Medioresinol	57.2	0.62	0.49	C21H24O7
MOL003184	Neotriptophenolide	45.42	0.53	0.85	C21H26O4
MOL003185	Triptonoterpenol	48.84	0.38	0.47	C21H30O4
MOL003187	Triptolide	51.29	0.68	0.25	C20H24O6
MOL003196	Tryptophenolide	48.5	0.44	1.11	C20H24O3
MOL003199	5,8-Dihydroxy-7-(4-hydroxy-5-methyl-coumarin-3)-coumarin	61.85	0.54	0.02	C19H12O7
MOL003217	Isoxanthohumol	56.81	0.39	0.76	C21H22O5
MOL003229	Triptinin B	34.73	0.32	0.84	C20H26O3
MOL003231	Triptoditerpenic acid B	40.02	0.36	0.97	C21H28O3
MOL003245	Triptonoditerpenic acid	42.56	0.39	0.81	C21H28O4
MOL003248	Triptonoterpene	48.57	0.28	1.22	C20H28O2
MOL003266	21-Hydroxy-30-norhopan-22-one	34.11	0.77	0.9	C29H48O2
MOL003280	Triptonolide	49.51	0.49	0.72	C20H22O4
MOL003283	Isolariciresinol	66.51	0.39	−0.2	C20H24O6
MOL005828	Nobiletin	61.67	0.52	1.05	C21H22O8
MOL007535	5alpha-Stigmastane-3,6-dione	33.12	0.79	0.9	C29H48O2
MOL009386	3,3′-bis-(3,4-dihydro-4-hydroxy-6-methoxy)-2H-1-benzopyran	52.11	0.54	0.14	C20H22O6

OB – oral bioavailability; DL – drug-likeness.

Screening for disease-related genes

We searched the GeneCards and OMIM databases using the keyword “nasopharyngeal carcinoma”, to obtain 1866 NPC related targets. Overlapping targets between the NPC-related target genes and the target genes of active ingredients were screened for. As a result, we identified a total of 30 overlapping genes between 1866 NPC related target genes and 62 target genes of active ingredients, as shown in Figure 1. After integrating and deleting duplicated genes, 22 active ingredients were identified.

Figure 1.

Venn diagram of disease related genes and drug targeted genes.

Drug-compound-target-disease network

We input 22 active ingredients and 30 “drug-disease” target genes into Cytoscape 3.6.1 software to construct a visualized drug-disease network. The network contained 54 nodes and 136 edges, as shown in Figure 2. A network analysis was conducted by assessing heterogeneity and centralization, which were found to be 1.131 and 0.491, respectively. Moreover, this network contained certain compounds with multiple targets, including high-degree compounds, such as MOL000422 (kaempferol, degree=19), MOL005828 (nobiletin, degree=12) and MOL000358 (beta-sitosterol, degree=9), as shown in Table 2.

Figure 2.

Drug-disease network analyses map.

Table 2.

Node degree of the drug–disease target network.

Name	Type	Degree	Name	Type	Degree
Nasopharyngeal carcinoma	Disease	30	MAPK8	Gene	3
Tripterygii Radix	Drug	22	CASP3	Gene	3
MOL000422	Mol	19	CASP9	Gene	3
NCOA2	Gene	17	RXRB	Gene	3
PGR	Gene	14	RELA	Gene	3
MOL005828	Mol	12	MOL009386	Mol	2
NCOA1	Gene	11	MOL007535	Mol	2
MOL000358	Mol	9	MOL002058	Mol	2
PPARG	Gene	7	MOL000211	Mol	2
MOL003283	Mol	6	MOL003266	Mol	2
ESR1	Gene	6	TIMP1	Gene	2
MOL003231	Mol	5	TP63	Gene	2
MOL003229	Mol	5	GSTM1	Gene	2
MOL000449	Mol	4	GSTP1	Gene	2
MOL003280	Mol	4	ALOX5	Gene	2
MOL003248	Mol	4	CYP1B1	Gene	2
MOL003217	Mol	4	VCAM1	Gene	2
MOL003196	Mol	4	SELE	Gene	2
MOL003187	Mol	4	ICAM1	Gene	2
MOL003184	Mol	4	CYP1A1	Gene	2
BCL2	Gene	4	CYP3A4	Gene	2
MOL003245	Mol	3	IKBKB	Gene	2
MOL003199	Mol	3	PON1	Gene	2
MOL003185	Mol	3	PRKCA	Gene	2
MOL000296	Mol	3	CASP8	Gene	2
CHEK1	Gene	3	CCR7	Gene	2
GSK3B	Gene	3	VEGFA	Gene	2

PPI network analysis

We constructed the PPI network (Figure 3) using the STRING database; 139 edges and 30 nodes were identified in the “drug-disease network” as shown in Figure 2. In order to better understand these genes based on the PPI network, we visualized the network using Cytoscape software and the plug-in Cytohubba to screen out hub genes with a degree of >16, which included estrogen receptor 1 (ESR1), vascular endothelial growth factor A (VEGFA), caspase 3 (CASP3) and RELA proto-oncogene, NF-κB subunit (RELA). Then, the top 4 hub genes with the highest degrees were identified (Figure 4).

Figure 3.

Protein target interaction network (PPI) of common target genes (the larger the size, the greater degree of the node).

Figure 4.

Protein interaction relationship histogram of common target genes.

Analyses of GO function and KEGG pathway enrichment

In order to deeply explore the functional role of the “drug-disease” target genes and key pathways involved in the use of Tripterygii Radix for the treatment of NPC, KEGG pathway enrichment and GO functional analyses were conducted (Figure 5). The results of the signaling pathway analysis indicated that these genes were mainly associated with pathways, including the PI3K-Akt signaling pathway, microRNAs in cancer, tumor necrosis factor (TNF) signaling pathway, and pathways in cancer, while the GO analysis indicated that these genes were significantly enriched in biological processes (BP), including DNA-templated, positive regulation of transcription, apoptotic process, interfering with transcription from RNA polymerase II promoter, response to drug, cell adhesion, and response to hypoxia.

Figure 5.

Enrichment analysis. (A) Enrichment analysis of Gene Ontology (GO) biological processes of common target genes. (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of common target genes.

Discussion

NPC was found to be associated with the Epstein-Barr virus (EBV), while non-keratinizing differentiated or undifferentiated carcinoma are predominant histological subtypes in endemic areas [27]. More than 70% of patients were diagnosed with locoregionally advanced disease at presentation [28]. According to the National Comprehensive Cancer Network (NCCN) guidelines, the combination of chemo-radiotherapy (CRT) is the standard treatment modality for locoregionally advanced NPC [29], although this treatment strategy may be promising, patients may suffer from a series of complications that significantly lower the quality of life. Additionally, considering the complex pathogenesis of NPC, single-target or single-drug treatment methods remain insufficient to exert a significant therapeutic effect. Therefore, it is essential to identify new therapeutic strategies and uncover its underlying molecular mechanism to provide new options for NPC patients.

Tripterygii Radix contains several active ingredients, which affect multiple targets and pathways of action, and it has been proven to be an inexpensive and effective method of treatment for many diseases. Triptonide, a component of Tripterygii Radix, exerts preventive potential against NPC [30]. Furthermore, Celastrol, which is also extracted from Tripterygii Radix, induces apoptosis of NPC cells through the ERK1/2 and p38 MAPK pathways [31]. Nevertheless, its pharmacological mechanisms of action in NPC treatment are still unclear. In the present study, the mechanism of action of Tripterygii Radix involved in the treatment of NPC was investigated using a network pharmacology approach. A total of 22 NPC related active compounds of Tripterygii Radix with an OB of ≥30% and a DL of ≥0.18 were selected using the TCMSP database, and the active compounds-target gene network was constructed. The results showed that the top 3 highest degree compounds were kaempferol, nobiletin, and beta-sitosterol. Kaempferol is a polyphenolic compound that has been described as a key element in inducing cancer cell apoptosis, as well as in suppressing angiogenesis and cancer cell growth [32]. Yoshida et al. showed that kaempferol could induce human colon cancer cell apoptosis by significantly upregulating TNF-related apoptosis-inducing ligand (TRAIL) receptors (DR5 and DR4) [33]. Nobiletin is a polymethoxy flavonoid, which was shown by Ma et al. to be able to significantly inhibit the growth of hepatic cancer cells via increasing the expressions of caspase-3 and Bax, while reducing the expressions of COX-2 and Bcl-2 [34]. Beta-sitosterol is a type of phytosterol, which was shown by Awad et al. to be able to significantly inhibit the growth of breast cancer cells by increasing caspase-8 activity [35].

The results of the PPI network analysis indicated that 4 hub genes, VEGFA, CASP3, ESR1, and RELA, were regulated by Tripterygii Radix in NPC. VEGFA is a significantly pro-angiogenic factor in cancer, which has been well studied during the past decades [36]. In NPC, Epstein-Bar virus induced VEGF was found to promote cancer metastasis through the recruitment and activation of macrophages [37]. High expression of VEGFA was found to be involved in the poor long-term survival of NPC patients [38]. CASP3 performs an essential function in executing cell apoptosis in tumor behaviors and is involved in both extrinsic and intrinsic cell death signaling pathways [39]. Chen et al. showed that genetic variation in CASP3 may contribute to increased risk of head and neck squamous cell carcinoma [40]. Abnormal expression of ESR1 often occurs in various of human epithelial cancers [41]. Marc et al. reported that hypermethylation ESR1 was associated with in EBV-positive NPC [42]. RelA (p65), a member of the NF-κB/Rel family, has been proven to be involved in drug resistance and migration of NPC [43,44].

In terms of GO enrichment, for our study we infer that Tripterygii Radix might exert its treatment effect by interfering with the positive regulation of response to drug, transcription, transcription of the RNA polymerase II promoter, apoptotic process, DNA-templated, cell adhesion, and response to hypoxia. RNA polymerase II is a type of RNAP enzymes, which is found in eukaryotic cell nucleus [45], and it has been linked to the growth of NPC cells [46]. Tan et al. showed that Tripterygii Radix could suppress the adhering of breast cancer cells by inducing the cleavage of focal adhesion kinase (FAK) [47]. Hypoxia-inducible factor-1α (HIF-1α), has been proven to be extensively linked with drug resistance, aggressive progression and tumor survival [48]. Tripterygii Radix may mediate this antitumor effect by reducing transcriptional activity and increasing the accumulation of hypoxia-inducible factor-1α [49].

The KEGG pathway enrichment analysis results showed an association with PI3K-Akt signaling pathway, microRNAs in cancer, TNF signaling pathway, and pathways in cancer. This is similar to that of previous studies on NPC molecular biology. For example, a number of studies have indicated the function of TNF-α in the inflammatory process related to carcinogenesis and the progression of cancer, while it also performs an essential function in enhancing angiogenesis and increasing the invasion and migration of tumor cells [50]. In the study by Yu et al., high expression of TNF-α was identified as an unfavorable prognostic indicator in NPC [51]. MiRNAs function as negative gene regulators and have been proven to inhibit critical cancer-related gene expressions and may be promising for the diagnosis and treatment of various cancers, including NPC [52]. The PI3K-Akt signaling pathway is closely associated with cancer proliferation, invasion, and metastasis and is one of the most common and significant signaling pathways for the progression of cancer [53]. Wang et al. showed that triptolide reduced the viability of NPC cells through the PI3K/Akt pathway [54]. Therefore, we proposed that the pharmacological effects of Tripterygii Radix in NPC may occur through these pathways.

This is the first study which included the systematical exploration of potential mechanisms of action of Tripterygii Radix in NPC. However, in vivo and in vitro experiments should be undertaken to validate the relationship between key genes and pathways of Tripterygii Radix for the treatment of NPC. Despite the limitations of this study, the results of this study provide new evidence and information to be used in subsequent theoretical and clinical research studies.

Conclusions

In conclusion, we utilized a systems pharmacology method by combining active ingredient screening, target prediction, PPI analysis, biological process, and KEGG pathway analyses to explore the fundamental molecular mechanism of action of Tripterygii Radix for its therapeutic effect in NPC. Some of these predicted targets and pathways are similar to the pharmacological effects reported in previous studies. The results of this analysis indicated that Tripterygii Radix acts on multiple targets and plays a therapeutic role in NPC through its action in multiple pathways.