Network Pharmacology-Based Strategy for Predicting Therapy Targets of Traditional Chinese Medicine Xihuang Pill on Liver Cancer

Objective To investigate the potential therapy targets and pharmacological mechanism of traditional Chinese medicine (TCM) Xihuang pill in liver cancer based on network pharmacology. Methods Drug ingredients-target network was constructed based on the target sets of Xihuang pill and liver cancer. The overlapping genes between Xihuang pill targets and liver cancer-related molecular targets were investigated using comparative analysis. Moreover, the PPI network and module was constructed based on overlapping genes and hub nodes, respectively, followed by the pathway enrichment analysis. Results A drug ingredients-target network was established with 1184 nodes and 11035 interactions. Moreover, a total of 106 overlapping genes were revealed between drug targets and liver cancer molecular targets. Furthermore, a PPI network and 4 modules were further investigated based on overlapping genes, respectively. These hub nodes such as VEGFA and EGFR were mainly enriched in GO functions including positive regulation of MAP kinase activity, activation of protein kinase activity, regulation of MAP kinase activity, and pathways like proteoglycans in cancer, bladder cancer, and estrogen signaling. Conclusion VEGFA and EGFR might be potential therapy targets of Xihuang pill in liver cancer. Furthermore, the effect of Xihuang pill on liver cancer might be realized by targeting VEGFA and EGFR in pathways like proteoglycans in cancer and estrogen signaling.


Introduction
As the second leading cause of death, liver cancer has caused a wide social burden for a long period of time [1]. Although partial surgical resection is the optimal therapy strategy for patients with liver cancer, the recurrence rates after surgery are still very high [2,3]. us, exploring the effective clinical treatment for liver cancer is necessary.
Traditional Chinese medicine (TCM) has been widely used for clinical treatment of various tumors such as liver cancer [4,5]. Xihuang pill is a complementary and alternative medicine that has been used in TCM due to the inhibition for tumor cell proliferation [6]. Xihuang pill is composed of Ru Xiang (olibanum), Mo Yao (Commiphora myrrha), She Xiang (Moschus), and Niu Huang (calculus bovis), which exert multiple antitumor effects [7]. A previous study shows that Xihuang pill has an anticancer effect on breast tumor [8], gastric cancer, and primary liver cancer [7]. Actually, Xihuang pill could improve quality of life and clinical manifestations of advanced primary liver cancer patients [9]. A previous study indicates that there is a reversal effect of serum containing Xihuang pill on the multidrug resistance of liver cancer cells via P-glytoprotein pathway [10]. Although Xihuang pill is effective in treatment of liver cancer, the drug targets and pharmacological mechanisms of Xihuang pill on liver cancer are still unclear. e network pharmacology analysis is a useful tool for further understanding of the drug action [11], which may provide insights into how we can improve drug discovery for complex diseases [12]. In the current study, the potential mechanism of Xihuang pill on the treatment of liver cancer was analyzed by a network-based systematic study (as shown in Supplementary Figure 1). Briefly, the drug ingredientstarget network was constricted based on the target sets of Xihuang pill and liver cancer. en, the overlapping genes were investigated using comparative analysis. Moreover, the networks of protein-protein interaction (PPI) and modules were constructed based on overlapping genes and hub nodes, followed by the GO and KEGG pathway enrichment analysis. We hoped to explore the potential therapy targets of Xihuang pill in the treatment of liver cancer, which can provide the basis for the pharmacological mechanism study of Xihuang pill.

Identification of Ingredients of Drug.
e chemical ingredients of Xihuang pill were collected from the Traditional Chinese Medicine Systems Pharmacology (TCMSP) Database (http://lsp.nwsuaf.edu.cn/tcmsp.php) [13] and the Traditional Chinese Medicine Integrated Database (TCMID, http://www.megabionet.org/tcmid/) [14]. According to the relevant parameters of the pharmacokinetic properties, the ingredients were screened according to oral bioavailability (OB) and drug-likeness (DL) values and those ingredients with DL ≥ 0.18 and OB ≥ 30% were selected as the active ingredients [15,16].

Drug Target.
Usually, the ingredients of drugs play a role in related biological functions via targets. To predict the targets of ingredients of Xihuang pill, the small molecular structure information of active ingredients in Xihuang pill was retrieved on the PubChem database (https://pubchem. ncbi.nlm.nih.gov/) [17]. Subsequently, the targets were screened with a screening online tool called the Swiss Target Prediction (http://www.swisstargetprediction.ch/) [18]. Finally, all the target information was standardized using UniProt (http://www.UniProt.org/).

Liver Cancer Targets.
e main source of disease targets for liver cancer was obtained from IPA (http://www. ingenuity.com). e targets of liver cancer were retrieved after deleting duplicate data. en, the target datasets were compared by Geneweaver (http://www.geneweaver.org) [20], and the overlapping targets of both liver cancer and the drug ingredients were considered potential targets of Xihuang pill for the treatment of liver cancer.

PPI Network Construction and Module Analysis.
According to STING database (http://www.string-db.org/) [21] with the species limited to "Homo sapiens" and the confidence score >0.9, the protein interaction pairs associated with overlapping target genes between Xihuang pill and liver cancer were screened. And the PPI network was constructed via utilizing the network visualization software Cytoscape (version: 3.0.0, http://chianti.ucsd.edu/cytoscape-3.0.0/) [11], and the topological properties of the PPI network were analyzed through NetworkAnalyzer (default settings). Furthermore, MCODE (version 1.5.1) [22] was used to screen the significant enriched modules from the PPI network.

Hub Gene Investigation in PPI Network.
e hub genes in PPI network were further investigated by a Cytoscape plugin cytoHubba [23], and 11 network topology parameters, including Degree, Edge Percolated Ingredients, Maximum Neighborhood Ingredients, Density of Maximum Neighborhood Ingredients, Maximal Clique Centrality Bottleneck, EcCentricity, Closeness, Radiality, Betweenness, and Stress were selected for importance calculation. After screening the core nodes in PPI network, the rank sum ratio (RSR) [24] was used to rank the nodes synthetically to get the core nodes of the network. e more important the output of a node, the higher the value of this node in the network.

Enrichment Analysis for the Hub Genes.
In order to reveal the potential biological function of Xihuang pill in the treatment of liver cancer, GO (Gene Ontology) functions [25] and KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway [26] enrichment analysis were performed. Specially, GO terms were grouped into the following three categories: biological process (BP), molecular function (MF), and cellular component (CC) [25]. And the KEGG analysis was conducted according to the enrichment analysis tool of the Database for Annotation, Visualization and Intergrated Discovery (DAVID, version: 6.8) software [27]. A p value <0.05 was considered as the cutoff criterion.

Drug Ingredients-Target Network Analysis.
A total of 53 ingredients and 1131 targets of Xihuang pill, as well as 566 molecular targets of liver cancer, were obtained in the current study. Based on these data, the drug ingredientstarget network was constructed. As shown in Figure 1, the network consists of 1184 nodes and 11035 interactions (see details in Supplementary File 1). Moreover, the top 50 nodes including 3 targets were further selected to construct the module of the drug ingredients-target network ( Figure 2; Supplementary File 2).

Overlapping Genes between Drug Targets and Liver Cancer Molecular Targets.
After the comparative analysis, the overlapping genes, which might be potential drug therapy target for Xihuang pill in liver cancer, between drug targets and liver cancer therapy targets were obtained. e result showed that a total of 106 overlapping genes (Attachment 1), including VEGFA, EGFR, ESR1, PLG, and MAPK3, were revealed in the current study.
Additionally, the KEGG pathway enrichment analysis showed that these overlapping genes were mainly enriched in pathways, such as metabolic pathways, pathways in cancer, proteoglycans in cancer, estrogensignaling pathway, and HIF-1-signaling pathway. e top 20 pathways enriched by overlapping genes were listed in Table 1.

PPI Network Analysis.
Based on the potential pharmacodynamic target of Xihuang pill for liver cancer, a PPI network was constructed by using the STRING. e result showed that there were 102 nodes and 766 interactions in the network ( Figure 3). Moreover, a total of 4 modules including module 1 (score � 9.913), module 2 (score � 2.143), module 3 (score � 1.5), and module 4 (score � 1.333) were further investigated from PPI network using MOCDE (Figures 4(a)-4(d)).

Hub Nodes Investigation.
According to the PPI network, the core nodes of the network were screened based on the characteristics of the network's topological structure ( Figures 5 and 6). e top 20 important nodes including VEGFA (RSR � 0.1218), EGFR (RSR � 0.1382), ESR1 (RSR � 0.1427), PLG (RSR � 0.1436), and MAPK3 (RSR � 0.1464) were summarized in Table 2. Moreover, the association among top 5 hub nodes, drug ingredients, and CTM was listed in Table 3.

Investigation of Pathways Enriched by Hub Nodes.
e top 5 hub nodes including VEGFA, EGFR, ESR1, PLG, and MAKP3 were used for further GO function and pathway enrichment analyses. e result showed that the hub genes could be assigned to different GO terms for BP, CC, and MF categories. e prominent functions enriched by hub genes were positive regulation of MAP kinase activity, activation of protein kinase activity, regulation of MAP kinase activity and phosphatidylinositol bisphosphate kinase activity, of which the top 10 functions for GO terms are presented in Figure 7. Moreover, these hub nodes were mainly enriched in pathways like proteoglycans in cancer (ID: 05205; count � 4; FDR � 8.93E-06), bladder cancer (ID: 05219; count � 3; FDR � 8.93E-06), and estrogen-signaling pathway (ID: 04915; count � 3; FDR � 6.48E-05). e top 10 pathways are listed in Table 4.

Discussion
Liver cancer is one of the most lethal cancers having worldwide prevalence, and the effect of clinical treatment for this disease is not satisfactory [28]. e current study, for the first time, explored the potential therapy targets of Xihuang pill on liver cancer based on network pharmacology analysis.
e results showed that a drug ingredients-target network was established with 1184 nodes and 11035 interactions.  Moreover, a total of 106 overlapping genes were revealed between drug targets and liver cancer molecular targets. Furthermore, a PPI network and 4 modules networks were further investigated based on overlapping genes and hub nodes, respectively. ese hub nodes such as VEGFA, EGFR, ESR1, PLG, and MAPK3 were mainly enriched in pathways like proteoglycans in cancer, bladder cancer, and estrogensignaling pathway. Vascular endothelial growth factor (VEGF) family plays a major role in angiogenesis, which are essential for both healing of injured tissue and proliferation of carcinoma cells [29]. e previous meta-analysis indicates that there is a prognostic role for VEFGA in various cancers including cervical cancer, gastric cancer, epithelial ovarian cancer, and liver cancer [30][31][32][33]. Taniguchi et al. indicates that VEGF promotes proliferation of hepatocytes through reconstruction of liver sinusoids by proliferation of sinusoidal endothelial cells [33]. Importantly, VEGF is an important drug target of various TCM [34,35]. A previous study shows that the TCM Fuzhenghuayu has a decoction effect on VEGF secretion in hepatic stellate cells [36]. However, the association between VEGF and Xihuang pill in liver cancer is unknown. Moreover, epidermal growth factor receptor (EGFR) is a transmembrane protein associated with the epidermal growth factor family [37]. e compartmentalization and biological function of EGFR in liver plasma membrane have been focused [38]. A previous study indicates that there is an increment of serum EGFR level in liver cancer patients [39]. Actually, there is a dual role of EGFR in liver injury and regeneration after acetaminophen overdose in mice [40]. Zhu et al. indicate that the EGFR status is independently correlated with TCM treatment in non-smallcell lung cancer patients [41]. Although previous study has confirmed the relation between EGFR expression and TCM active compounds using cell membrane chromatography [42], the effect of Xihuang pill on EGFR expression in liver cancer is still unclear. In the current study, VEGFA and EGFR were not only overlapping genes revealed between Xihuang pill target sets and liver cancer target sets but also hub nodes in PPI modules. us, we speculated that genes such as VEGFA and EGFR might be potential therapy targets of Xihuang pill in liver cancer. Proteoglycans perform multiple functions in cancer and angiogenesis due to their polyhedric nature [43]. e deterioration of liver function is accompanied by an increase in the amount of chondroitin sulfate proteoglycans [44]. e   Figure 6: e results of hub nodes in current PPI network. e ranking of hub nodes based on degree. From red to yellow represented a decline in importance; the line between two nodes represented the interaction.      Evidence-Based Complementary and Alternative Medicine alteration of proteoglycan composition interferes with the physiologic function of the liver on several levels [45]. Actually, the biological function of proteoglycans has been widely investigated in VEGF-induced diseases [46,47]. A previous study shows that the activation of EGFR contributes to the inhibition of axon regeneration [48]. Furthermore, estrogen plays a vital role in the progression of liver cancer [49]. Kim et al. indicate that estrogen-related receptor γ is upregulated in liver cancer [50]. Previous studies show that VEGF and EGFR expression are all associated with estrogen receptor status in patients with cancer [51,52]. Meanwhile, cytoplasmic expression of estrogen receptor can predict poor outcome of EGFR-TKI therapy in metastatic lung adenocarcinoma [53]. In this study, the pathway analysis showed that the potential targets such as VEGFA and EGFR were mainly enriched in pathways including proteoglycans in cancer and estrogen-signaling pathway. us, we speculated that the effect of Xihuang pill on liver cancer might be realized by targeting VEGFA and EGFR in pathways like proteoglycans in cancer and estrogen signaling. However, there were some limitations in this study such as small sample size and lack of verification analysis. us, a further verification study based on a large sample size is needed.

Evidence-Based Complementary and Alternative Medicine
In conclusion, VEGFA and EGFR might be potential therapy targets of Xihuang pill in liver cancer. Furthermore, the effect of Xihuang pill on liver cancer might be realized by targeting VEGFA and EGFR in pathways like proteoglycans in cancer and estrogen signaling.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Additional Points
Highlights. (1) VEGFA and EGFR might be potential therapy targets of Xihuang pill in liver cancer. (2) Proteoglycans in cancer and estrogen signaling might be associated with liver cancer. (3) Totally, 106 genes were explored between drug targets and liver cancer targets

Conflicts of Interest
e authors declare that they have no conflicts of interest.