Network Pharmacology-Based Study to Uncover Potential Pharmacological Mechanisms of Korean Thistle (Cirsium japonicum var. maackii (Maxim.) Matsum.) Flower against Cancer

Cirsium japonicum var. maackii (Maxim.) Matsum. or Korean thistle flower is a herbal plant used to treat tumors in Korean folk remedies, but its essential bioactives and pharmacological mechanisms against cancer have remained unexplored. This study identified the main compounds(s) and mechanism(s) of the C. maackii flower against cancer via network pharmacology. The bioactives from the C. maackii flower were revealed by gas chromatography-mass spectrum (GC-MS), and SwissADME evaluated their physicochemical properties. Next, target(s) associated with the obtained bioactives or cancer-related targets were retrieved by public databases, and the Venn diagram selected the overlapping targets. The networks between overlapping targets and bioactives were visualized, constructed, and analyzed by RPackage. Finally, we implemented a molecular docking test (MDT) to explore key target(s) and compound(s) on AutoDockVina and LigPlot+. GC-MS detected a total of 34 bioactives and all were accepted by Lipinski’s rules and therefore classified as drug-like compounds (DLCs). A total of 597 bioactive-related targets and 4245 cancer-related targets were identified from public databases. The final 51 overlapping targets were selected between the bioactive targets network and cancer-related targets. With Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment, a total of 20 signaling pathways were manifested, and a hub signaling pathway (PI3K-Akt signaling pathway), a key target (Akt1), and a key compound (Urs-12-en-24-oic acid, 3-oxo, methyl ester) were selected among the 20 signaling pathways via MDT. Overall, Urs-12-en-24-oic acid, 3-oxo, methyl ester from the C. maackii flower has potent anti-cancer efficacy by inactivating Akt1 on the PI3K-Akt signaling pathway.


Introduction
The definition of cancer is that normal cells are damaged via an aberrant endogenous process such as an abnormality during DNA replication or instability of the DNA sequence, which transforms into malignancy [1]. The DNA damage responses represent chronic inflammation via the immune signaling pathway, which results in accelerating tumorigenesis [2]. The damaged normal cells undergo cellular senescence, triggering secretion in the inflammatory cytokines leading to cellular mechanical disruption [3,4]. Because inflammation is a leading factor in causing pathological symptoms such as unknown severe pain, fatigue, and comorbidity in cancer patients, anti-inflammation strategies are thus key therapeutics [5,6]. Current anti-inflammatory agents for cancer treatment NSAIDs, including Cox-2, due to fewer adverse effects and a lower mortality rate [7]. Most commonly, anti-cancer agents aim to inhibit DNA replication and induce cancer cell death; however, cancer chemotherapeutics also attack healthy cells resulting in serious side effects such as nausea, vomiting, hair loss, and fatigue [8][9][10]. On the other hand, traditional herbal plants with innovative bioactives and secondary metabolites play an essential role as effective anti-inflammatory, anti-oxidant, or anti-cancer agents [11]. For instance, plant extracts (Urtica membranaceae, Artemesia monosperma, and Origanum dayi Post) in combination with anti-cancer drugs showed enhanced potency against specific cancer cell lines (lung, breast, colon, and prostate cancer) without exposing normal lymphocytes and fibroblasts to cytotoxicity [12]. Herbal-derived bioactives possess fewer unwanted side effects than chemotherapy, and have led to new clinical drugs such as taxol from Taxus brevifolia L., vincristine from Catharanthus roseus G. Don, and Epipodophyllotoxin from Podophyllum peltatum L. [12,13].
C. maackii is a perennial herbal plant, belonging to the family of Compositae, and is widely distributed in the mountainous areas of Korea, Japan, and China [14]. Furthermore, Cirsium species have been reported to have diverse pharmacological activities such as antioxidant, anti-inflammation, anti-cancer, and hepatoprotection effects [15][16][17][18]. Specifically, a C. maackii extract at a concentration of 200 µg/mL showed 36.89% inhibition against a breast cancer cell line (MDA-MB-231) [19]. Another study demonstrated that HepG2 cells treated with a MeOH extract of C. maackii have potent antioxidant efficacy against severe oxidant conditions [20]. Anti-inflammatory agents assist in protection against cancer development, thereby preventing cytokine storms [21]. It implies that anti-inflammatory compounds are important agonists to protect normal cells adjacent to tumor cells because they can block the overflow of cytokines. Until now, C. maackii flower compounds were identified by HPLC and had only been reported for anti-Alzheimer efficacy by inhibiting BACE1 [22]. Generally, the identification of polar and mid-polar compounds from extracts is based on HPLC due to its good separation capability [23,24]. From a different perspective, we utilized GC-MS analysis to discover lipophilic bioactives, which mainly act as drug-like compounds and uptake efficiently into the cells. Lipophilicity is a significant physicochemical parameter that influences membrane permeability and affinity [25]. More importantly, GC-MS, along with the molecular docking test (MDT) and ADME (Absorption, Distribution, Metabolism, and Excretion) study, is an optimal analytical method to determine drug-like compounds [26]. At present, the bioactives and mechanisms of the C. maackii flower against cancer remain unknown. Hence, we aimed to uncover its potential bioactives with their fundamental mechanisms through network pharmacology.
Network pharmacology (NP) as a systemic method can analyze holistic bioactivetarget-disease relationships [27]. It can decipher the unknown mechanism(s) with "multiple targets", "multiple bioactives", instead of "one target", "one bioactive" [28]. This approach is a very effective method to recognize the mechanism of action for lead compounds discovered from herbal plants [29]. An existing drug may be re-modelled to bind on multiple targets with the concept of NP, thus it can be a guide for drug repurposing [30]. The NP application is a powerful tool to elucidate novel targets and bioactives from natural products, and is especially effective for anti-cancer research to investigate multi-target activity in biological pathways and interaction networks [31]. Here, we implemented NP to provide key bioactive(s), uppermost target(s), and potential mechanism(s) of the C. maackii flower against cancer. Mainly, the flower part of herbal plants are the most underexplored part compared to other parts such as leaves, roots, and stems. During the growing season of the C. maackii plant, the flowering part might be utilized as a source of essential bioactives or be taken as a functional food. In this study, we suggest that C. maackii flowers are valuable parts with potential anti-cancer compounds.
To prove their therapeutic value, we performed a GC-MS analysis, protein network investigation, and a molecular docking test (MDT). Its brief processes are discussed below.
Firstly, bioactives from the C. maackii flower were identified by GC-MS analysis and screened to find drug-likeness compounds via an in silico tool. Then, targets associated with bioactives or cancer were identified through public databases, and final overlapping targets were utilized to analyze protein-protein interaction (PPI) and the highest degree of a target. Thirdly, a signaling pathway-target protein-bioactive (S-T-B) relationship against cancer was identified by networking analysis. Lastly, we found the most potent bioactive and a hub target to alleviate cancer severity by exploring the molecular mechanism of the C. maackii flower based on MDT. The workflow diagram is displayed in Figure 1.

Bioactives from C. maackii Flower
A total of 34 bioactives in the C. maackii flower were identified by GC-MS analysis (Figure 2), and the name of compounds, PubChem ID, retention time (mins), and peak area (%) are listed in Table 1. All 34 compounds were accepted by Lipinski's rules (Molecular Weight ≤ 500 g/mol; Moriguchi octanol-water partition coefficient ≤ 4.15; Number of Nitrogen or Oxygen ≤ 10; Number of NH or OH ≤ 5), and all bioactives corresponded with the standard of "Abbott Bioavailability Score (>0.1)" through SwissADME. The TPSA (Topological Polar Surface Area) value of all bioactives was also accepted ( Table 2).  Protective cell against pathogens [44] PCIDB: PhytoChemical Interactions DB.

Overlapping Targets between SEA and STP Associated with Bioactives
A total of 309 targets from SEA and 396 targets from STP connected to 34 bioactives were identified (Supplementary Table S1). The Venn diagram showed that 108 targets overlapped between the two public databases (Supplementary Table S1) ( Figure 3A).

Overlapping Targets between Cancer-Associated Targets and the Final 51 Overlapping Targets
A total of 4245 targets related to cancer were selected via retrieval from TTD and OMIM databases (Supplementary Table S2). The Venn diagram results revealed 51 overlapping targets that were selected between 4245 targets associated with cancer and 108 overlapping targets ( Figure 3B) (Supplementary Table S3).

Acquisition of a Hub Target from PPI Networks
From STRING analysis, 46 out of 51 overlapping targets were directly related to cancer occurrence and development, indicating 46 nodes and 145 edges ( Figure 4). The five targets removed (PAM, EPHX1, PPARD, KDM5C, and BCHE) had no connectivity to the overlapping 51 targets. In PPI networks, the Akt1 target was the highest degree (29) and was considered a hub target (Table 3).

Identification of a Hub Signaling from Bubble Chart
The output of KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway enrichment analysis revealed that 20 signaling pathways were associated with 24 targets (False Discovery Rate < 0.05). The 20 signaling pathways were directly connected to cancer, suggesting that these 20 signaling pathways might be the noteworthy pathways of the C. maackii flower against cancer. The descriptions of 20 signaling pathways are displayed in Table 4. Additionally, a bubble chart suggested that the PI3K-Akt signaling pathway might be a hub signaling pathway of the C. maackii flower against cancer ( Figure 5). Among the 20 signaling pathways, the Akt1 target was associated with 18 signaling pathways, representing the highest degree of value. Most importantly, Akt1 is directly related to the PI3K-Akt signaling pathway whereas both the PPAR signaling pathway and the Calcium signaling pathway are not correlated with Akt1.

Toxicological Properties of a Selected Key Bioactive
Additionally, the toxicological properties of Urs-12-en-24-oic acid, 3-oxo-, methyl ester were predicted by the admetSAR online tool. Our result suggested that the bioactive did not disclose Ames toxicity, carcinogenic properties, acute oral toxicity, or rat acute toxicity properties (Table 10). Table 10. Toxicological properties of the uppermost bioactive on AKT1 (PDB ID: 5KCV) in MDT.

Discussion
The S-T-B network suggested that the therapeutic efficacy of the C. maackii flower against cancer was directly associated with 20 signaling pathways, 24 targets, and 19 bioactives. Through the network, we identified the most significant protein (Akt1) associated with the occurrence and development of cancer and a bioactive (Urs-12-en-24-oic acid, 3-oxo-, methyl ester) from the C. maackii flower. From a bubble chart, we identified a hub signaling pathway (PI3K-Akt signaling pathway) connected to the Akt1 target, indicating the lowest rich factor among 20 signaling pathways. A report demonstrated that the activated PI3K-Akt signaling pathway accelerates tumor cell proliferation, invasion, and metastasis, inhibiting apoptosis [46]. Furthermore, it was found that the Akt1 target was overexpressed in 15 out of 24 human hepatocellular carcinomas (63.3%) confirmed using PCR analysis through Northern blot [47]. Another research study suggested that the PI3K-Akt signaling pathway's antagonists are potential anti-cancer candidates to regulate acute and chronic inflammatory responses [48].
In the S-T-B network, Urs-12-en-24-oic acid, 3-oxo-, methyl ester had the highest degree of value and was considered the uppermost bioactive of the C. maackii flower against cancer. Urs-12-en-24-oic acid, 3-oxo-, methyl ester is categorized into boswellic acids used widely as anti-inflammatory agents, including as an anti-cancer treatment. Additionally, the boswellic acids have potent anti-cancer efficacy against diverse malignant cancers [49,50]. The KEGG pathway enrichment analysis of 24 targets suggested that a total of 20 signaling pathways were involved in cancer occurrence and development. The relationships of the 20 signaling pathways with cancer are concisely discussed as follows. In the Peroxisome Proliferator-Activated Receptor (PPAR) signaling pathway, the activation of the PPAR signaling pathway functions as an anti-inflammatory agent, which can overwhelm the metabolic energy balance of cancer cells by inhibiting the fatty acid synthesis and accelerating fatty acid oxidation [51,52]. In the Mitogen-Activated Protein Kinase (MAPK) signaling pathway, MAPK inhibitors are efficient blockers to reduce pro-inflammatory cytokines and enhance the anti-cancer effect, particularly on human pancreatic cancer cells [53,54]. In the Rap1 (Ras-associated protein-1) signaling pathway, Rap1 promotes cytokine production during the inflammatory condition, which leads to tumor progression in human colorectal cancer [55,56]. Regarding the alcium signaling pathway, calcium is a significant second messenger to regulate inflammation; hence, blockers of the calcium channel can induce cancer cell death [57,58]. In the Cyclic AMP (cAMP) signaling pathway, an increased cAMP level has an anti-inflammatory effect, where the increased level can regulate DNA damage, DNA repair, and apoptosis of cancer cells [59,60]. Concerning the Hypoxia-Inducible Factor-1 (HIF-1) signaling pathway, HIF-1 is a central regulator to stimulate the production of inflammation. HIF-1 overexpression is related to increased tumor growth [61,62]. In the sphingolipid signaling pathway, sphingolipid is implicated in the inflammatory response, which has been involved in cancer cell proliferation [63,64]. In the Phospholipase D (PLD) signaling pathway, PLD inhibition can induce two functions, namely anti-inflammation and anti-cancer. Mainly, the blocking of PLD during chemotherapy can sensitize one to chemotherapeutics [65,66]. Regarding the Phosphoinositide 3-kinase-Akt (PI3K-Akt) signaling pathway, inhibition of the PI3K-Akt signaling pathway reduces the severity of inflammation in mice. This strategy is a promising mechanism for the treatment of cancers such as lung cancer, colorectal cancer, renal cancer, prostate cancer, triple-negative breast cancer, mucinous adenocarcinoma of the ovary, and skin cancer [67,68]. In the AMP-activated protein kinase (AMPK) signaling pathway, AMPK activation suppresses inflammatory responses and dampens cancer growth with cell metabolism and the cell cycle [69,70]. Regarding the Vascular Endothelial Growth Factor (VEGF) signaling pathway, VEGF is a core mediator in the formation of new blood vessels for cancer cells; thereby, cancer cells can survive, grow, and metastasize [71]. In the B cell receptor (BCR) signaling pathway, B cells are involved in the inflammatory T cell receptor, which is implicated in antibody production [72]. B cells or BCR-associated kinases may function with anti-cancer activity via B cells activation [73]. In the Fc epsilon RI signaling pathway, the expression of Fc epsilon on mast cells stimulates immunoglobulin E, leading to type 1 hypersensitivity-induced local inflammatory responses at the tumor sites [74]. In the insulin signaling pathway, insulin-resistant patients undergo excessive production of reactive oxygen species (ROS) that can harm DNA attributed to carcinogenesis [75]. In the estrogen signaling pathway, estrogen exposure to chronic inflammatory disease activity is a key risk in breast cancer progression [76]. With regards to the prolactin signaling pathway, prolactin functions as a cytokine immune system, especially in breast cancer, which has the strongest correlation with an increased expression level of prolactin and prolactin receptors [77,78]. In the thyroid signaling pathway, the optimal regulation of the cellular thyroid hormone is essential for an adequate role of immune cells during inflammation [79]. In terms of the adipocytokine signaling pathway, adipocytokine is an inflammatory mediator during immune-associated diseases, which accelerates cancer progression and metastasizes from organ to organ [80]. In the Relaxin signaling pathway, Relaxin alleviates the inflammatory severity and diminishes the amount of leucocytes and the expression level of cytokines [81]. In terms of the Advanced Glycation End-product (AGE)-Advanced Glycation End-product Receptor (RAGE) signaling pathway in diabetic complications, RAGE can stimulate inflammatory responses by binding with AGEs [82]. The inhibitor (papaverine) of RAGE is a promising target for anti-cancer activity, which blocks nuclear factor kappa B (NF-κB) [83].

Plant Material Collection and Identification
The C. maackii flowers were collected from Mihogil of Bomunmyeon (Latitude: 36

Plant Preparation, Extraction
The C. maackii flower was dried in a shady area at room temperature (20-22 • C) for 7 days, and dried leaves were powdered using an electric blender (Shinil, Cheonan, Korea). Approximately 50 g of C. maackii flower powder was soaked in 800 mL of 100% methanol (Daejung, Seohaean, Korea) for 5 days and repeated 3 times to collect extraction. The solvent extract was collected, filtered, and evaporated using a vacuum evaporator (RV8, IKA, Staufen, German). The evaporated sample was dried under a boiling water bath (HB10, IKA, Staufen, German) at 40 • C to obtain the extract.

GC-MS Analysis Condition
Agilent 7890A (Agilent, Santa Clara, CA, USA) was used to carry out the GC-MS analysis. GC was equipped with a DB-5 (30 m × 0.25 mm × 0.25 µm) capillary column (Agilent, Santa Clara, CA, USA). Initially, the instrument was maintained at a temperature of 100 • C for 2.1 min. The temperature rose to 300 • C at a rate of 25 • C/min and was maintained for 20 min. The injection port temperature and helium flow rate were sustained at 250 • C and 1.5 mL/min, respectively. The ionization voltage was 70 eV. The samples were injected in the split mode at 10:1. The MS scan range was set at 35-900 (m/z). The fragmentation patterns of mass spectra were compared with those stored in the W8N05ST Library MS database. The percentage of each compound was calculated from the relative peak area of each compound in the chromatogram. The concept of integration was used with the ChemStation integrator (Agilent, Santa Clara, CA, USA) algorithms (analyzed 19 May 2021) [84].

Bioactives Database Construction and Drug-Likeness Property
The bioactives from the C. maackii flower were identified by utilizing GC-MS analysis. Then, the GC-MS-detected bioactives were filtered in accordance with Lipinski's rules through SwissADME (http://www.swissadme.ch/) (accessed on 3 June 2021) to confirm the "Drug-likeness" physicochemical properties. PubChem (https://pubchem.ncbi.nlm. nih.gov/) (accessed on 3 June 2021) was utilized to select the SMILES (Simplified Molecular Input Line Entry System) bioactives.

Construction of PPI Networks and Bubble Chart
For the final overlapping targets, STRING (https://string-db.org/) (accessed on 16 June 2021) [87] was utilized to analyze the PPI network. Thereby, RPackage was used to identify the degree of value. Then, signaling pathways on STRING were visualized by RPackage, a hub signaling pathway (lowest rich factor) related to a hub target (highest degree of value from PPI).

Construction of a Size Map on S-T-B Network
Both the hub target (the highest degree of value among 20 signaling pathways) and the uppermost bioactive were identified via the S-T-B network. The S-T-B networks were utilized to construct a size map, based on the degree of values. In this size map, green rectangles (nodes) represented signaling pathways, pink triangles (nodes) represented targets, and orange circles (nodes) represented bioactives. The circle size represented the degree value, the size of pink triangles represented the number of connectivity with signaling pathways, and the size of orange circles represented the number of connections with targets. The merged networks were constructed using RPackage.

Preparation for MDT of Ligand Molecules
The ligand molecules were converted from sdf in PubChem into the .pdb format using Pymol, and the ligand molecules were converted into the .pdbqt format through Autodock.

Ligand-Protein Docking
The ligand molecules were docked with targets utilizing autodock4 by setting up an energy range of 4 and exhaustiveness at 8 as the default to obtain 10 different positions of ligand molecules [89]. The center (the position of the middle coordinate point) in the target was X: −12.677, Y: 2.931, Z: −13.145 on AKT1 (PDB ID: 5KCV). The grid box size was set to 40 Å × 40 Å × 40 Å. The 2D binding interactions were used with LigPlot+ v.2.2 (https://www.ebi.ac.uk/thornton-srv/software/LigPlus/) (accessed on 23 June 2021) [90]. After docking, ligands of the lowest binding energy (highest affinity) were selected to visualize the ligand-protein interaction in Pymol.

Toxicological Properties Prediction by admetSAR
Toxicological properties of the key bioactive were established using the admetSAR web-service tool (http://lmmd.ecust.edu.cn/admetsar1/predict/) (accessed on 24 June 2021) [91] because toxicity is an essential factor in developing new drugs. Hence, Ames toxicity, carcinogenic properties, acute oral toxicity, and rat acute toxicity were predicted by admetSAR.

Conclusions
The bioactives and mechanisms of C. maackii flowers against cancer were firstly uncovered through network pharmacology. The findings suggested that 20 signaling pathways, 24 targets, and 19 bioactives are connected to cancer. Of these, the PI3K-Akt signaling pathway, Akt1, and Urs-12-en-24-oic acid, 3-oxo-, methyl ester were the hub signaling pathway, hub target, and key bioactive of C. maackii flowers against cancer, respectively. Furthermore, Urs-12-en-24-oic acid, 3-oxo-, methyl ester has the most potent efficacy on the Akt1 target protein than 13 other standard ligands. This study suggests that the mechanism of the C. maackii flower against cancer might strengthen anti-inflammatory responses by inactivating the PI3K-Akt signaling pathway, bound to Urs-12-en-24-oic acid, 3-oxo-, methyl ester on Akt1. From this viewpoint, we propose that C. maackii flowers can be utilized as functional or medicinal resources against cancer.
Supplementary Materials: The following are available online, Table S1: The 309, 396, and 108 targets from SEA, STP, and overlapping targets between SEA and STP, respectively.  Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data generated or analyzed during this study are included in this published article (and its Supplementary Information Files).
Acknowledgments: This research was acknowledged by the Department of Bio-Health Convergence, College of Biomedical Science, Kangwon National University, Chuncheon 24341, Korea.

Conflicts of Interest:
There is no conflict of interest.
Sample Availability: Samples of described compounds are available from the authors.