Network pharmacology combined with molecular docking and molecular dynamics to verify the therapeutic potential of mung beans (Vigna radiata) against prostate cancer

Abstract

Background

Prostate cancer is the most common oncological disease in men and one of leading causes of death worldwide. Growing evidence has demonstrated the effectiveness of mung bean bioactive compounds in suppressing various cancer cells. However, their effects and underlying mechanisms on prostate cancer have not been verified. The present study aimed to investigate the therapeutical effects and underlying mechanisms of mung bean compounds against prostate cancer.

Results

The results revealed that 56 proteins related to prostate cancer could be modulated by mung bean, including several vital proteins of SRC (Sarcoma), Mitogen-Activated Protein Kinase 8 (MAPK8), Heat shock protein 90 kDa alpha member A1 (HSP90AA1), and Harvey Rat sarcoma virus (HRAS). It was also found that the potential pathways associated with prostate cancer pathogenesis comprising pyrimidine metabolism, nitrogen metabolism, and prolactin signaling pathways. Of 19 mung bean compounds docked to four key proteins reveal three promising compound (dulcinoside, peonidin-3-glucoside, and chlorogenic acid) with lower binding affinity score of − 7.7, − 12.2, − 9.0, and − 6.5 kcal/mol against SRC, MAPK8, HSP90AA1, and HRAS, respectively in their site of action. Dynamic simulation results also showed values of − 36.52 ± 2.93, − 35.93 ± 1.67, and − 35.77 ± 1.17 kJ/mol for Dulcinoside-SRC, Dulcinoside-MAPK8, and P3G-HSP90AA1 complexes, respectively. The binding of the compound occur in stable and flexible with the proteins. Moreover, all mung bean compounds predicted to have good ADMET properties.

Conclusions

The study concluded that dulcinoside, peonidin-3-glucoside, and chlorogenic acid potentially exhibited anticancer activity against prostate cancer in silico. Nevertheless, further studies such as in vitro and in vivo are needed to optimize and prove the efficacy of the mung brand and its compounds against prostate cancer.

Graphical abstract

1 Background

Prostate cancer is the second most commonly diagnosed cancer in men worldwide after lung cancer. It constitutes significant public health issue as the fifth most frequent cause of cancer-related deaths globally. The various risk factors for prostate cancer are consumption of foods rich in animal saturated fats, red meat, and dairy products, environment, excessive intake of alcohol and coffee, and vitamin D deficiency [1, 2]. The clinical presentation of prostate cancer is associated with decline in the quality of life including sexual, physical, and psychosocial domains [3, 4]. Moreover, the risk of developing cardiovascular diseases and suicide is elevated following a diagnosis of prostate cancer [5].

Bioactive compounds that exert as antioxidants, anti-inflammatory, antilipidemic, and anticancer agents are widely recognized for their beneficial effects on human health. These compounds are predominantly found in fruits and vegetables [6, 7]. In particular, mung bean (Vigna radiata) compounds have emerged as promising candidate for anticancer treatments. These plants have demonstrated the ability to confer chemoprotection against breast and cancer cells in vitro and in vivo [8]. The same ability has also been exposed against hepatocarcinoma cancer cells in vitro and in vivo in female BALB/C mice (inbred mice with ability to produce monoclonal antibody) [8].

Mung beans are rich source of flavonoid and phenolic acids, extensively studied for their potential health benefits. Flavonoid compounds of mung bean include anthocyanins, flavanols, flavones, flavonols, and isoflavonoids [9]. In vitro studies have exhibited that anthocyanins can inhibit cell proliferation in the HT29 colon cancer cell line [10]. Flavanol-derived compounds have been found to affect the cell cycle by modulating pathways such as factor-κB, mitogen-activated kinase protein, epidermal growth factor, vascular endothelial growth factor, and matrix metalloproteinase [11, 12]. Besides, flavonol-derived compounds have also been discovered to have anticancer effects by modulating the caspase pathway-3, Bax, and Bcl to induce apoptosis in the PA-1 cell line [7, 11, 13, 14]. Meanwhile, flavone-derived compounds have been investigated for their antineoplastic effects on cell lines of various cancer types, including breast cancer (MCF-7), leukemia (U937), brain tumor (PC12), and esophageal cancer (EC-109) [15, 16].

Phenolic acid compounds in mung beans are hydroxycinnamic acid and hydroxybenzoic acid derivatives and have various health benefits including anticancer effects. Some hydroxycinnamic acid-derived compounds have been uncovered to have cytotoxic effect on cancer cell lines and modulate apoptosis activity [17,18,19]. Hydroxybenzoic acid derivative compounds have been shown to have similar products to hydroxycinnamic acid derivatives. It modulates the apoptosis process by targeting several proteins such as Bcl-2, caspase-3, and caspase-9. These findings suggest that mung beans can be an anticancer agent [19, 20]. Nevertheless, further research is required to explore these compounds' potential in vivo and better understand their mechanisms.

Despite the ongoing research on prostate cancer, the effects and underlying mechanisms of polyphenol derivates of mung bean on this disease are still inadequately investigated. To uncover, several comprehensive approaches can be adopted such as network pharmacology combined with molecular docking. Besides, molecular dynamics can be considered since these methods have been widely utilized in drug discovery against complex diseases like cancer.

Network pharmacology is scientific method that can help to discover protein-related prostate cancer targeted by mung bean compounds. It is done by merging the fields of bioinformatics, pharmacology, biology, and computer science. Network pharmacology is an innovative analytical approach representing a paradigm shift from the traditional "one-target, one drug" approach to the "network-target, multiple-component therapeutics" approach. It has been considered to assess the effects and underlying mechanisms of diseases since it can provide information on multiple biological processes, metabolic pathways, and drug/compound-target interactions [21, 22].

Furthermore, molecular docking is an established technique that analyzes interactions between drug-like ligands and protein target receptors using computational algorithms. This technique helps to identify suitable active sites, obtain the best geometry, and calculate ligand interaction energies for more effective compound development. Moreover, it can be used to verify the interaction of mung bean polyphenolic compounds with essential prostate cancer proteins and underlined residue and type of binding interaction [23, 24].

Consequently, network pharmacology and molecular docking can be integrated to validate the former and identify affinity bindings. Additionally, molecular dynamic simulations can be used to analyze the interaction stability and flexibility of complexes formed between the active compound of mung bean and critical proteins related to prostate cancer.

Accordingly, this study aimed to analyze the effects and underlying mechanisms of polyphenolic compounds of mung bean on prostate cancer using network pharmacology combined with molecular docking and molecular dynamics. The workflow of the study is presented in Fig. 1.

Fig. 1

Workflow of network pharmacology combine with molecular docking and dynamic stimulation to verify therapeutic potential of mung bean against prostate cancer

2 Methods

2.1 Retrieval of polyphenol compounds of mung beans

The literature search revealed 19 different polyphenol compounds of mung bean (Vigna radiata) that were included in this study. They were cyanidin-3-glucoside, peonidin-3-glucoside, pelargonidin-3-glucoside, quercetin, myricetin, kaempferol, catechin, vitexin, isovitexin, luteolin, dulcinoside, p-coumaric acid, caffeic acid, ferulic acid, chlorogenic acid, and sinapic acid, gallic acid, syringic, and gentisic acid [8, 9, 25].

2.2 Analysis of protein related polyphenol compounds of mung bean and Prostate cancer

The targets of polyphenol compounds mung bean were determine using two integrative online based tools, SwissTargetPrediction and PharmMapper by employing SMILES and SDF file of the compounds. SwissTargetPrediction was set in default mode. Meanwhile, PharmMapper was adjusted to “HumanOnly” and the selection targets were conduct by z-score value of > 1.5. The retrieved protein were merged and the duplicate is removed.

Furthermore, the searching for protein-related prostate cancer was conducted using two different databases, including GeneCards and Online Mendelian Inheritance in Man (OMIM) with the keyword "Prostate Cancer". The results of protein analysis were merged and the duplicate is removed.[26].

Afterward, the collected target of mung bean compounds were analyzed and intersected with prostate cancer related protein using Venny 2.1.0. Identified overlapped protein assumed as core target of mung bean in prostate cancer [27].

2.3 Construction of protein–protein interaction network

The STRING database were used to construct protein–protein interactions (PPI) network for intersecting proteins in previous section. To ensure robustness results, the screening threshold of the tools were set in confidence level > 0.9 with “Homo Sapiens” mode. Furthermore, the PPI was visualized and analyzed in Cytoscape 3.9.1 and the free proteins (proteins that are not bound to the main network) were removed [27].

2.4 Analysis of topological network and screening of key targets

Analysis of topological protein was computed by utilizing the cytoscape plug in, CytoNCA. In this analysis, we use four different centrality including degree (DC), betweenness (BC), eigenvector (EC), and closeness centrality (CC). The higher node score indicated to have a crucial role within the PPI network. Furthermore, The main target protein is analyzed from the 10 proteins with the highest value in each of the four centrality (DC, BC, EC and CC) which then intersected in ven diagram to obtain the main protein [28].

2.5 Gene ontology enrichment and KEGG signal pathway analysis

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation were performed by ShinyGO v 0.77 with specified organism “Homo Sapiens”. GO and KEGG pathway analysis was conducted to examine the protein cluster within the network in their influence on enriched biological process (BP), molecular function (MF), cellular component (CC) and pathway related prostate cancer. The cut-off FDR were set below 0.05 to obtain the top 10 GO annotation and KEGG pathrway lollipop plot [29].

2.6 Molecular docking analysis

Molecular docking analysis was conducted to assess the ability of the mung bean compounds to bind with screened key proteins of prostate cancer. The docking was carried out by AutoDock Vina program package. The selected compound of mung bean as well as Docetaxel and Bicalutamide (as positif control drug of prostate cancer) were retrieved from the PubChem database and further prepare using Marvin View and autodock into.pdbqt format. Meanwhile, the key protein including SRC (PDBID: 1O4N), MAPK8 (PDBID: 4L7F), HSP90AA1 (PDBID: 1OSF), and HRAS (PDBID: 4DLS), were obtained from Research Collaboratory for Structural Bioinformation Protein Data Bank (RCSB PDB). The protein wre further prepare using BIOVIA Discovery Studio Visualizer and autodock to obtaind.pdbqt format of the proteins. Prior docking, we also determine the grid box parameters to ensure docking of compound with selected protein occure in catalytic site. All docking were conduct by AutoDock Tools 1.5.6 and the results were visualized by BIOVIA Discovery Studio Visualizer 2021 to display the 2D and 3D structure [28]

2.7 Prediction of ADMET and drug likeliness profiles of mung bean compounds

The ADMET prediction was conducted to determine absorption, distribution, metabolism, excretion and toxicity of the compounds. The prediction was completed by pkCSM (https://biosig.lab.uq.edu.au/pkcsm/), SwissADME (http://www.swissadme.ch/index. php), and admetSAR (http://lmmd.ecust.edu.cn/admetsar2/).

2.8 Dynamic simulation and calculation of

Molecular dynamics of proteins with ligands were performed via OpenMM and AMBER force fields run in Google Colab. The temperature was set constant at 310 K. and a time step of 2 fs.

3 Result

3.1 Potential targets of mung bean compounds in prostate cancer

The potential targets of mung bean (Vigna radiata) compounds were analyzed using PharmMapper and SwissPrediction tools. After assessing the database and removing duplicate targets, 211 proteins targeted by mung bean compounds were obtained. Meanwhile, after searching GeneCard and OMIM databases and removing repeated proteins, 7.528 prostate cancer-related proteins were achieved. Furthermore, both groups of proteins (drug and disease) were intersected using Venn diagrams (Venny 2.1.0). It provided 159 overlapping proteins that were considered potential targets of mung bean compounds in prostate cancer therapy (Fig. 2).

Fig. 2

Venn Diagram Analysis Intersection of Prostate cancer related protein with mung bean compound target

4 Network contruction, topological analysis and key protein screening

Next, 159 prostate cancer proteins targeted by mung bean compounds were further examined to determine the interaction of these proteins network and topological analysis. The protein protein interaction were buid using online based tools STRING and Cytoscape software package. The results of PPI displayed in Fig. 3a indicated that some of the proteins had no interaction with others and those in the main cluster network that further were removed using Cytoscape. The final network results 56 nodes proteins with 121 edges protein protein interaction (Fig. 3b). Furthermore, analysis of topological network were conduct using CytoNCA (cytoscape plug in) provide different score in four different centrality including degree (DC), betweenness (BC), eigenvector (EC), and closeness centrality (CC). The higher node score indicated to have a crucial role within the PPI network (Table 1). Menggunakan plug in dan centrality yang sama, key protein screening dilakukan untuk memperoleh putative and most potential protein in prostate cancer that targeted by mung bean compound. Top 10 protein pada masing masing sentrality were retrieved and further intersected by venn diagram (Fig. 4A–D). This analysis revealed four potential proteins in the interaction such as SRC, MAPK8, HSP90AA1, and HRAS and suggested as key protein (4E). The proteins obtained were further constructed on the Drug–Compound–Target–Disease Network to make the final visualization in the form of network pharmacology (Fig. 5).

Fig. 3

Protein–Protein Interaction network of mung bean compound target in prostate cancer. A The PPI Network of 159 protein target of mung bean compound target in prostate cancer therapy according to STRING. B The PPI Network of 56 protein target of mung bean compound target in prostate cancer therapy after clearing non interaction protein with main network

Table 1 Topological of protein rated prostate cancer targeted by mung bean polyphenol compounds

Fig. 4

Top 10 potential target protein in different centrality based on highest score. A Betweenness B Closeness C Degree D Eigenvector E Venn diagram of top 10 proteins that figurize four obtained key proteins including SRC, MAPK8, HSP90AA1, and HRAS

Fig. 5

Drug–Compound–Target–Disease Network with 4 potential key protein

4.1 GO enrichment and KEGG signal pathway analysis

Gene Ontology analysis was performed to analyze molecular function, cellular components, and biological processes associated with prostate cancer targeted by mung bean compounds. It aimed to predict comprehensive picture of the changes in molecular functions, cellular components, and biological processes that occur in prostate cancer when mung bean compounds were challenged into the disease. There were 468 molecular functions and several essential ones found, namely glutathione transferase activity and MAP kinase activity. Meanwhile, 200 cellular components related to the input protein were also uncovered. Some important cellular features included the ficolin-1-rich granule lumen and endosome lumen. Another finding was that the proteins involved 1000 biological processes, including the glutathione derivative metabolic process and cyclooxygenase pathway. KEGG was conducted to analyze the pathway mainly enriched in prostate cancer-related proteins targeted by mung bean compounds. In this analysis, 206 pathways involved in 56 proteins inputted were uncovered. Several critical protein-related pathways comprise prostate cancer, pyrimidine metabolism, nitrogen metabolism, and prolactin signaling pathways. The top 10 items were then ranked based on the number of annotations to functional area (Fig. 6a–d).

Fig. 6

Functional annotation and KEGG pathway enrichment analysis of 56 core prostate cancer targets of mung bean compound. A Top 10 of Biological Process. B Top 10 of cellular component. C Top 10 of molecular function. D Top 10 of KEGG pathways

4.2 Molecular docking

The 19 compounds were docked with primary potential targets, including SRC, MAPK8, HSP90AA1, and HRAS, to verify the therapeutic potential of mung mean compounds. They included cyanidin-3-glucoside, peonidin-3-glucoside, pelargonidin-3-glucoside, quercetin, myricetin, kaempferol, catechin, vitexin, isovitexin, luteolin, dulcinoside, p-coumaric acid, caffeic acid, ferulic acid, chlorogenic acid, and sinapic acid, gallic acid, syringic, and gentisic acid. The validation was performed using molecular docking. It is shown that all compound could interact with the proteins. Dulcinoside become compound with the lowest docking score against SRC targets (∆G = − 7.7 kcal/mol) and MAPK8 (∆G = − 12.2 kcal/mol) (Fig. 7). Meanwhile, P3G possessed the lowest value against the HSP90AA1 target (∆G = − 9.0 kcal/mol) and luteolin on HRAS (∆G = −7.2 kcal/mol) (Fig. 8 and Table 2.) Although luteolin was docked on a non-active site, chlorogenic acid was chosen for further investigation of residue interactions. In addition, the three compounds (dulcinoside, peonidin-3-glucoside, and chlorogenic acid) involved had better binding values than the controls. Furthermore, as presented in Table 3, several interactions occurred in the compounds and target bonds, namely hydrogen bonds, Van der walls, Pi–Pi T-Shaped/Pi–Sigma, Unfavorable bump, Carbon Hydrogen Bond, Pi–Sulfur/Sulfur–X interaction, Pi–Cation, Alkyl/Pi–Alkyl Interaction, and Halogen bonds.

Fig. 7

Visualization of 2D and 3D of molecular docking results of best protein–ligand (mung bean compound and positive control) complexes A. SRC-Dulcinoside B SRC-Dasatinib (control) C MAPK8-Dulcinoside D MAPK8-Pyrazolanthrone (control)

Fig. 8

Visualization of 2D and 3D of molecular docking results of best protein–ligand (mung bean compound and positive control) complexes A HSP90AA1-P3G B HSP90AA1-Geldanamycin (control) C HRAS-ChlorogenicAcid D HRAS-Kobe0065 (control)

Table 2 The docking result of key protein target with mung bean compound and positive control

Table 3 Interaction residue key protein target with mung bean compound and positive control

4.3 Prediction of ADMET and drug likeliness profiles of mung bean compounds

ADMET prediction were essential aspects of drug discovery. In this study, three different tools including pkCSM, SwissADME, and admetSAR, were employed since they can accurately predict the ADME and toxicity of the compounds/drugs. The results exhibited that the values of intestinal absorption of the test compounds were relatively lower than that of the positive control. Although the skin permeability of all test compounds had a more negative value of logKp than the control, they had poor skin permeability compared to the control. Each test compound was not permeable to Caco2 and had no P-gp substrate or inhibitor. The Volume Distribution steady state (VDss) values of all test compounds had better deals, and only peonidin-3-glucoside had subcellular localization in the nucleus, while others did in the mitochondria (Table 4). Compared to the positive control, all test compounds were poor in blood–brain barrier permeability. The test compounds could not inhibit cytochrome and were not a substrate of the cytochrome. In addition, none of the compounds that predicted hepatotoxic, carcinogenic, and toxic to the AMES model (Table 5). All three compounds violate Lipinski's rule. Dulcinoside had a significant molecular weight, hydrogen acceptor of > 10, and hydrogen donor of > 5. Meanwhile, peonidin had a hydrogen acceptor of > 10 and a hydrogen donor of > 5. Finally, chlorogenic acid had hydrogen donor > 5 (Table 6).

Table 4 Pharmacokinetics Properties of 3 selected mung bean compound and corresponding positive control

Table 5 Toxicity Properties of 3 selected mung bean compound and corresponding positive control

Table 6 Druglikeness of 3 selected mung bean compound

4.4 Molecular dynamic and binding free energy

The stability and dynamic of ligand–protein interaction in top ligand–protein complexes were further observed. Molecular dynamics were simulated for 5 ns to analyze root mean square deviation (RMSD), root mean square fluctuation (RMSF), and Radius of Gyration (RoG). Based on Fig. 9a–d, the RMSD values of three compound-receptor complexes, including Dulcinoside-SRC, Dulcinoside-MAPK8, and P3G-HSP90AA1, fluctuated and were less than 2.5 Å. It indicates that the binding was relatively stable. Based on Fig. 8a, the RMSD of the Dulcinoside-SRC interaction was stable at 3 ns initially, then increased at 3.5 ns, and finally stable at 4–5 ns, indicating it reached equilibrium. Meanwhile, the equilibrium phase of the Dulcinoside-MAPK8 interaction was predicted in 3–5 ns. In contrast to the three complexes, the chlorogenic acid-HRAS interaction demonstrated abnormal fluctuations with RMSD values ranging from 0.5 to 6 Å.

Fig. 9

RMSD, RMSF and Radius of Gyration of selected protein–ligand complexes during molecular dynamic stimulation A Dulcinoside-SRC B Dulcinoside-MAPK8 C Peonidin-3-Glucoside-HSP90AA1 D Chlorogenic acid-HRAS

Additionally, in line with RMSD, the RMSF values of three compound-receptor complexes, including Dulcinoside with SRC, Dulcinoside with MAPK8, and Peonidin-3-glucoside with HSP90AA1, demonstrated good fluctuations and indications of stable interactions since the values ranging from 0.5 to 6.0 Å. With the exception of chlorogenic acid-HRAS, the 170 amino acid residue obtained an RMSF value of about 30 Å. Other outputs from the molecular dynamic (Fig. 9a–d) were RoG values ranging from 13.05–13.40 Å (Dulcinoside-SRC), 22.1–22.7 Å (Dulcinoside-MAPK8), 16.9–17.4 Å (Peonidin-3-glucoside -HSP90AA1), and 14.8–15.3 Å (chlorogenic acid-HRAS). RoG value measures how compact the protein is with the ligand molecule. The figures show that the gyrase values were stable, and no sudden change in RoG values existed. The values indicate that the binding of protein with ligand occurs very compactly.

Furthermore, we also found that the average binding free energy (ΔG) of three receptors-ligand complexes, including Dulcinoside-SRC, Dulcinoside-MAPK8, and P3G-HSP90AA1, exhibited almost the same values of − 36.5187 ± 2.93, − 35.93 ± 1.67, and − 35.7723 ± 1.17 kJ/mol, respectively. Meanwhile, the chlorogenic acid-HRAS interaction produced a binding free energy of − 12.5533 ± 1.65 kJ/mol (Table 7).

Table 7 Free energy binding of selected protein-mung bean compound complexes during molecular dynamic stimulation

5 Discussion

Recently, drug discovery and development have entered new era with the concept of going back to nature. The concept is realized through natural product-based compound approach that has been proven effective in curing various diseases, including prostate cancer. Among these, several compounds have been suggested to provide anticancer activity against prostate cancer [30]. Mung bean contains many bioactive compounds that can potentially be used as cancer drugs, such as flavonoid, alkaloid, and tannin derivatives compounds [8, 9, 25].

This study comprehensively explained the effects of the mung bean compounds on prostate cancer and the essential proteins and pathways by applying a network pharmacology approach. Furthermore, this study also verified the underlying mechanisms using molecular docking dynamics and investigated the binding free energy with the MM/GBSA approach.

Network pharmacology was conducted to understand the interaction of each compound with its biological targets to provide information on potential targets that play role in prostate cancer progression [22]. The analysis found that there were 56 main protein targets associated with prostate cancer. Further network analysis specifically uncovered four key potential targets, including SRC, MAPK8, HSP90AA1, and HRAS. Conversely, functional enrichment analysis discovered the 56 proteins associated with various biological, molecular, and pathway processes linked to prostate cancer. As a result, GO enrichment found that these proteins could involve molecular function in prostate cancer. Two essential pathways in the pathogenesis of prostate cancer comprised glutathione transferase activity and MAPK activity. The MAPK pathway contributes to prostate cancer progression, where p38, c-Jun N-terminal kinases (JNK), and Extracellular signal-regulated kinase (ERK) proteins play an essential role in cell survival, apoptosis, and cell differentiation [31]. Furthermore, Glutathione Transferase P1 (GSTP1) can be marker of prostate gland carcinogenesis, whereas methylation of GSTP1 is an epigenetics associated with prostate cancer [32].

Furthermore, in cellular component analysis (Fig. 6B), two cellular components involving 56 prostate cancer proteins targeted by the mung bean compounds were highlighted, namely ficolin-1-rich granule lumen and endosome lumen. The ficolin-1-rich granule is a cellular component commonly used for prostate cancer metastasis [33, 34]. In addition, endosome involvement can also be one of the biomarkers for prostate cancer diagnosis. Research by Johnson et al. exposed that there is a specific disconnect between the initial cellular endosome (peripheral cells) location and the late cellular endosome (perinuclear cells). This specific disconnect can affect the degradation and signaling processes in prostate cancer cells [35].

Moreover, molecular processes that occur in the body of an organism will affect its biological functions. In this analysis, several biological processes were associated with prostate cancer incidence, namely the glutathione derivative metabolic process and cyclooxygenase pathway. Glutathione and its related parts are essential in tumor initiation, development, and drug resistance. Glutathione is synthesized in the cytosol and plays a vital role in preventing the detrimental effects of reactive oxygen species (ROS) on mitochondria in the electron transport process. Glutathione works do not depend on the amount of antioxidants in cancer cells. In prostate cancer tissue, the circulation of glutathione peroxidase is significantly reduced. Thus, it can change the intracellular environment into a prooxidant state and cause significant changes in gene expression that can lead to malignancy [36, 37]. Meanwhile, the cyclooxygenase pathway, especially cyclooxygenase-2 (COX-2), becomes one that plays a role in the course of prostate cancer. COX-2 is reported to be overexpressed and tends to be elevated in prostate cancer. Studies involving prostate cancer cell line PC-3 and LNCaP suggest upregulation of expression of COX-2 mRNA and increased cell proliferation [38].

Several enrichment pathways in prostate cancer proteins were targeted by mung bean compounds, including nitrogen metabolism, pyrimidine metabolism, and prolactin signaling pathways. All three pathways are associated with the incidence of prostate cancer. Increased nitrogen requirements are considered one of the essential metabolic features of cancer cells. This is due to the ability to maintain proliferative signals in cancer. Cell proliferation must synthesize nucleotides containing essential nitrogen [39]. In prostate cancer, amino acid metabolism plays a vital role in the development of cancer cells. One thing that plays a function is the precursors of nitrogen-containing metabolites, such as purines and pyrimidines, for nucleic acid synthesis [40]. In addition, an increase in the rate of nitrogen metabolism occurs along with an improvement in the metabolic rate of glutamate and aspartate in prostate cancer [41]. Pyrimidines are a component of nucleotides in the process of cell proliferation. The presence of disturbances in the pyrimidine metabolic process is associated with the progression of cancer, including prostate cancer. Research conducted by Kelly et al. revealed pyrimidine metabolism and oxidative phosphorylation were the most dysregulated pathways in the lethal type of prostate tumor (p < 0.007) [42]. Furthermore, the hormone Prolactin (PRL) can also play a role in cell proliferation, survival, and tumorigenesis of prostate cancer cells. Suppression of the hormone prolactin can be a consideration in the treatment of prostate cancer [43, 44].

Silico docking and dynamic analysis of 19 polyphenol compounds against four critical proteins, including SRC, MAPK8, HSP90AA1, and HRAS, reveal promising interaction. The results demonstrated that the three best compounds for each protein target, showing the lowest average binding free energy in catalytic site, were dulcinoside, peonidin-3-glucoside, and chlorogenic acid. The smaller binding free energy indicates a better binding affinity of the compound to its protein target. Compounds derived from mung beans had a lower affinity value compared to the control compounds [45]. Therefore, it can be assumed that these compounds have better inhibitory potential than the controls.

The Src protein is one of the proto-oncogenes that play a role in signal transduction during cellular activities, such as cell differentiation, adhesion, and cell migration. This protein plays a role in androgen-dependent and androgen-independent stages of prostate cancer [46]. In vitro, researchers conveyed that the inhibition of this protein can be helpful in the treatment of prostate cancer. The inhibition of Src using Dasatinib has been tested in prostate cancer. The result revealed that Dasatinib could become a suppressor agent for cancer cells and significantly reduce the incidence of lymph gland metastases [47].

The MAPK8/JNK1 protein can act as a proapoptotic agent while also inducing cell proliferation, invasion, and migration [48]. JNK1 ATP-competitive inhibitor, Pyrazolantrhone (SP600125), is less specific than Betamapimod (AS602801). The inhibition of MAPK8/JNK1 protein in prostate cancer cells using JNK inhibitors and enzalutamide can affect cell death, inhibit the proliferation, migration, and invasion of prostate cancer cells, and prevent cell growth. There is opposition to the process of JNK1 inhibition due to the benefits of inhibiting cell proliferation, invasion, and metastasis while inhibiting the function of apoptosis [49].

HSP90AA1 is one of the protein subtypes of the HSP90 family. This protein is located in the cytosol. The HSP90 protein is involved in cellular processes and regulates apoptotic pathways, cell cycles, and signaling. HSP90 protein may promote prostate cancer progression in the Nuclear factor kappa B (NF-kB) pathway. This protein can also regulate the process of prostate cancer proliferation and apoptosis through many pathways such as receptor pathways androgen, human epidermal growth factor receptor 2 (ERBB2), Act, c-RAF, survivin, Epidermal growth factor receptor (EGFR), Insulin-like growth factor 1 (IGFR-1), Signal transducer and activator of transcription 3 (STAT3), ERK, Cyclin-dependent kinase 4 and 6 (CDK-4 and CDK-6) signaling pathways. HSP90 inhibitors as a therapy against cancer can use Geldanamycin, a potent antitumor activity, but this compound has an unstable structure and is hepatotoxic [50].

HRAS is a protein from the RAS family that can contribute to tumorigenesis, invasion, and metastasis of various types of cancer. Inhibiting this protein can be a treatment option to prevent cancer cell proliferation, aggression, and migration [51]. Furthermore, Kobe0065 family compounds can inhibit the interaction of Ras-GTP with many effectors, including RAF, Phosphoinositide 3-kinase (PI3K), Ral guanine nucleotide dissociation stimulator (RalGDS), and Son of Sevenless (SoS). This results in the inhibition of the cellular activity of the HRAS pathway [52]. In addition, the use of Simvastatin in inhibiting Cav1 may decrease the expression of the H-RAS/(PLCε) pathway, which is known to hinder migration Castration Resistant Prostate Cancer (CRPC) [53].

The ADMET profiles of drugs/candidates are essential to drug discovery. Assessment of absorption, distribution, metabolism, excretion, and toxicity is critical to demonstrate. In this study, the absorption component was reviewed by assessing the parameters of intestinal absorption, Caco2 permeability, skin permeation, and P-glycoprotein substrate and inhibitor. The intestinal absorption component can predict the proportion of compounds absorbed through the human small intestine, where a value of < 30% indicates a poor absorption rate. According to Table 4, several compounds had a good absorption rate. Only dulcinoside had a value slightly lower than the threshold, so it can be predicted that this compound will not be absorbed appropriately by the intestine. [54].

Caco2 permeability is one component that can be employed to predict drug absorption when administered orally. It can be done because this model can express cytochrome P450 enzymes, transporters, microvilli, and enterocytes identical to the human small intestine [55]. All test compounds could not penetrate Caco2. Skin permeation is a component that can predict if a drug can penetrate the skin. All test compounds had relatively lower skin permeation values compared to the controls, so it can be said that they are poorly permeable to the skin [56]. Furthermore, they also showed no potential substrate or inhibitor for P-glycoprotein (P-gp). Hence, P-gp could not actively transport them, and the efflux activity of P-gp was reduced [57, 58].

In this study, distribution parameters in the test components were Volume Distribution steady state (VDss) and Blood blood–brain barrier (BBB) permeation. VDss is a component that can predict the total dose of the drug distributed in tissues. It is considered low if the VDss log value is < − 0.15, while high if the value is > 0.45. Based on Table 4, all test compounds had high VDss log values, so these compounds are predicted to have good network distribution capabilities [54, 59]. Further, BBB permeation refers to the ability of a compound to be permeable to the blood–brain barrier. The results revealed that all test compounds are predicted to have poor permeability to BBB [60]. Therefore, these compounds are not suitable when used in cases of prostate cancer that metastasizes to the brain. However, these compounds need to be retested in vitro or in vivo to assess the permeability quality of BBB compounds. In addition, a substrate such as BBB permeabilizer kinin analogs can be added to increase the permeability of the BBB so that the compound can pass through the BBB and affect its biological target [61].

Furthermore, metabolic predictions in this test included the ability to inhibit CYP2D6, CYP3A4, CYP1A2, CYP2C19, and CYP2C9 and substrates of CYP2D6 and CYP3A4, which are the amino acid residues. All test enzymes are part of the cytochrome P450 family, which is very important clinically and plays a role in drug metabolism [54, 59, 62,63,64]. Several anticancer drugs often must be metabolized by cytochrome P450 enzymes to become active or be excreted from the body, such as tamoxifen, an antiestrogen drug for treating breast cancer [65].

All test compounds could not inhibit all test enzymes and were not substrates of CYP2D6 and CYP3A4. Hence, it can be concluded that the compounds will not impede the work and will not be metabolized by cytochrome P450. Besides metabolism, components excrete compounds in the body by assessing the approximate total cleansing log (CLtot). The total clearance value can be applied as a reference in determining the dose of the drug and understanding the mechanism by which the drug is removed from the body [66, 67]. The last parameter tested was toxicity, while the components evaluated were max tolerated dose, AMES toxicity, Carcinogens, hepatoxicity, Tetrahymena Pyriformis toxicity, Acute oral toxicity, oral rat acute toxicity (LD50), and oral rat chronic toxicity (LOAEL). Tetrahymena Pyriformis toxicity value demonstrates the total dose of the molecule for the inhibition of T. Pyriformis by 50% of growth. The LD50 value refers to the dose of a substance killing 50% of the tested sample. Meanwhile, the LOAEL value denotes the minimum amount of a sense that can have side effects when consumed in the long term. LD50 and LOAEL values can be employed as a reference to determine safe and effective drug doses and assess potential harm to the organism [68, 69].

AMES toxicity is commonly used to understand better and predict DNA mutation affected by a given chemical [70]. At the same time, carcinogens are conducted to assess the chemical ability to induce carcinogenesis [70], and hepatotoxicity is tested to predict the chemical ability to be toxic to the liver [54]. In this study, it is found that all test compounds potentially did not have toxicity to the AMES model, did not possess carcinogens, and were toxic to the liver. The analysis of the drug-likeness aspect using the Lipinski of 5 rules assesses molecular weight parameters (< 500), hydrogen donors (< 5), hydrogen acceptors (< 10), LogP (< 5) or MLogP (< 4,15), and Molar Refractivity (40–130) [71]. The molecular weight parameter can indicate the ability of a compound to be absorbed through the wall of the small intestine. Hydrogen donor and hydrogen acceptor refer to the ability of compounds to interact and be soluble in water. Furthermore, the MLog P parameter is an indicator of lipophilicity. The higher the MLog P value, the more lipophilic a compound is. Last, molar refractivity refers to the ability of a compound to interact appropriately with its biological targets [72, 73].

An orally active compound should not have more than one violation. If a compound violates more than one parameter, the gastrointestinal tract is difficult to absorb, and its bioavailability is low. Despite this, many drugs violate Lipinski's rules but are still orally active. Lipinski's rule predicts a compound with the ability to diffuse passively. This rule is less relevant if a compound is a substrate transporter so that it can actively diffuse [72, 73]. Therefore, further pharmacokinetic properties testing should be performed to assess the ADMET component of the tested compound.

Based on this analysis, test compounds with > 1 violation of Lipinski's rule should not be taken orally. Still, a method of delivery is created so that the drug reaches its biological target, such as intravenously, intramuscularly, buccal, or anally [74]. In addition, it can use a nanoparticle as a delivery so that complex compounds can be easily absorbed and enter the cell to have biological effects on its target. Furthermore, these compounds can still be administered orally with the help of some absorption-enhancing components such as chitosan, surfactants, bile salts, nano-carrier, nano-emulsion, and dendrimers [75].

6 Conclusion

In summary, it can be concluded that the potential proteins related to prostate cancer from network pharmacology analysis were SRC, MAPK8, HSP90AA1, and HRAS. KEGG analysis revealed 206 mechanism pathways potentially associated with prostate cancer pathogenesis, including nitrogen metabolism, pyrimidine metabolism, and prolactin signaling pathway. Molecular docking results indicated that the test compounds had binding free energy values better than their controls. The compact and stable interaction assumed that the three best compounds (dulcinoside, peonidin-3-glucoside, and chlorogenic acid) had a better effect than drug control. Furthermore, the analysis revealed that the compounds are predicted to have a good pharmacokinetics and toxicology profile. However, studies should be conducted to verify the effectivity of mung bean compounds in prostate cancer, such as in vitro and in vivo studies.