Curcumin and capsaicin play a vital role in anti-inflammatory and anti-cancer mechanism as they are used as therapeutic drugs/adjuvants. Our previous study including many reports explored strong inhibitory effect of curcumin and capsaicin on lipopolysaccharide-induced polymorph blood mononuclear cells (PBMCs) and cancer cells. Therefore, a systematic study was carried out to identify the potential protein targets of curcumin and capsaicin in cancer as well as angiogenesis through network pharmacology and molecular docking approaches. In the present investigation, we employed integrative prediction of cancer targets of curcumin and capsaicin through the ChEMBL and STITCH databases, followed by network construction, network topology, gene ontology, pathway enrichment and molecular docking studies. The gene ontology analysis made it possible to identify a library of possible cancer targets of curcumin (34 targets) and capsaicin (35 targets). Based on topological analysis, the unique target of curcumin and capsaicin was proposed by identifying essential bottleneck/hub node MAPK1. Further, PANTHER gene set analysis was used to distinguish the biologically enriched pathways in top identified gene clusters (MAPK1). To validate the identified target, high-throughput molecular docking was employed as both molecules curcumin and capsaicin along with standard ulixertinib were docked against MAPK1 to understand the binding interaction. The docking results of MAPK1 with curcumin (− 7.6 kcal/mol) has shown good inhibitory effect similar to that of standard control ulixertinib (− 8.1 kcal/mol) compared with capsaicin (− 6.0 kcal/mol). Based on the molecular interaction, MAPK1 identified through the network pharmacology approach could be a probable target of curcumin and capsaicin to prevent angiogenesis in cancer.
Turmeric and red pepper have long been used as traditional medicine to treat burning sensations and as analgesics (Kempaiah et al. 2005; Manjunatha and Srinivasan 2006). The active spice principles of turmeric (curcumin) and red pepper (capsaicin) are known to exhibit anti-inflammatory, anti-carcinogenic, anti-metastatic, anti-angiogenic and chemopreventive properties by inducing apoptosis (Hatcher et al. 2008; Oyagbemi et al. 2010). Studies have confirmed that these ingredients are found to have anti-cancer properties (Kunnumakkara et al. 2008; Pyun et al. 2008; Sung et al. 2012). These studies have prompted us to further understand the molecular mechanism of curcumin and capsaicin at the molecular level by employing network pharmacology approach to obtain drug targets.
Network pharmacology is an emerging area that uses computational knowledge to develop an understanding of drug action across multiple scales of complexity ranging from single drug target to multiple drug targets. By integrating multi-faceted approaches, network pharmacology can offer mechanistic action of drugs on target proteins. The network pharmacology approach mainly aims at paradigm shift from “one drug, one target” to “multicomponent therapeutics/biological complexes” (Tang and Aittokallio 2014; Zhang et al. 2016; Xu et al. 2018). Thus, network pharmacology greatly enhances our knowledge by exploring the underlying mechanism of naturally extracted drugs and its targets. This contains studying how drugs can act on different cell types and tissues, as well as multiple actions within a distinct cell type by virtue of several cross-talk pathways. The biological process in the body can be represented in different ways. In general, protein–protein interaction complexes (PPI)/signaling pathways are the set of genes/proteins that interact with each other to achieve biological process, molecular function as well as cellular components involved in the set of genes/proteins. In KEGG (Kyoto Encyclopaedia of Genes and Genomes) signaling pathways proteins/genes are represented by nodes and signals/connections as edges during activation/repression of pathways that transmit signals from one gene to another (Kanehisa and Goto 2000). In large-scale free PPI network, an in silico validation and analysis of biological networks play a vital role in identifying hub nodes. The hub nodes are the well-connected functional proteins maintaining the global network and cell–cell communication in the complex biological system; these nodes are linked with each other via physical interactions (Raman 2010; Ran et al. 2013). The topological parameters emphasize the role of node in the interactome and GO analysis is used to evaluate protein clusters involved in the biological process of certain diseases (Sun et al. 2011; Liu et al. 2015), divulging the role of nodes in signaling pathways. To explore the drug targets, we employed data curation of genes/proteins associated with curcumin and capsaicin for constructing an interactive network to identify the drug targets to prevent angiogenesis during cancer.
Materials and methodology
Data acquisition and processing
For the identification of target genes/proteins associated with curcumin and capsaicin, a comprehensive literature survey was carried out using ChEMBL database (https://www.ebi.ac.uk/chembl/#) and STICH 5.0 database (http://stitch.embl.de/), which form part of EMBL-EBI used to manually curate the drug–gene interactive information (Gaulton et al. 2012; Szklarczyk et al. 2016). The data were subjected to functional enrichment analysis to acquire drug–gene interactions, which are essential for the core interactome construction and analysis, indicating the signaling pathways/candidate genes of cancer phenotype needed to be sorted out through “Database for Annotation, Visualization and Integrated Discovery” (DAVID), a functional enrichment analysis tool (https://david.ncifcrf.gov/) (Huang et al. 2007) to understand the biological significance which ultimately defines the role of cancer associated with curcumin and capsaicin gene targets. The candidate drug targets were further enriched using the KEGG (http://www.genome.jp/kegg/pathway.html) pathways to construct the drug targets—pathway relationship using the online tool STRING v.10 database, which aims to provide critical predicted information on protein–protein interactions derived from genomic resources, experimental evidences, text mining, co-occurrence and co-expression (Szklarczyk et al. 2015). The data in STRING are weighted and integrated by assigning a high confidence score of 0.7, which indicates the probability that the interaction is authentic. The enriched signaling pathways with false positive rate discovery (FDR) < 0.05 were used in the following study.
Gene ontology (GO) analysis
To examine the functional association of angiogenic targets of curcumin and capsaicin, we performed gene ontology (GO) enrichment analysis using PANTHER gene list analysis, which is used to evaluate the characteristics of genes associated with angiogenesis (http://www.pantherdb.org/) (Mi et al. 2016). On the basis of the information obtained from GO, cancer interactome was constructed and analyzed to identify the possible angiogenic hub targets of curcumin and capsaicin.
Network construction and module identification
The PPIs network of candidate genes associated with the curcumin and capsaicin molecule was constructed and visualized using open-source Cytoscape v 3.2.1 software (Shannon et al., 2003). To explore the important hub targets responsible in regulating the subnetwork, the bottleneck method of cytoHubba was used (Chin et al. 2014) to find the top ten genes with the shortest path considered as key regulators of the subnetwork. The scale-free evaluation of subnetwork was investigated and analyzed using Network Analyzer v.3.3.1 by power law curve fit (Assenov et al. 2008; Belenahalli Shekarappa et al. 2017).
To examine the minimum number of significant genes associated with specific biological process, molecular function and cellular components, the first-degree neighbors of hub protein were extracted and subjected to PANTHER GO analysis using the following default parameters (Mi et al. 2016).
Molecular docking study of the top hub node
The top hub/bottleneck protein MAPK1 identified through topology analysis was subjected to in silico molecular docking study. The X-ray crystallographic structure of MAPK1 (PDB ID: 1WZY, resolution: 2.5 Å) was retrieved from RCSB PDB (http://www.rcsb.org/pdb/home/home.do) (Rose et al. 2015). The hetero atoms and water molecules in the protein structure were removed using MGL tools, followed by the addition of essential hydrogen atoms, Gasteiger charges and torsion degrees of freedom. Solvation parameters were assigned and saved in PDBQT format for further analysis (Trott and Olson 2010). Ligand molecules curcumin (PubChem ID: 969516), capsaicin (PubChem ID: 1548943) and standard ulixertinib (PubChem ID: 11719003) were downloaded from the PubChem server (https://pubchem.ncbi.nlm.nih.gov/) (Wang et al. 2009) and structural geometry optimization was carried out using the PRODRG server (http://davapc1.bioch.dundee.ac.uk/cgi-bin/prodrg) (vanAalten et al. 1996; Kandagalla et al. 2017).
AutoDock Vina (ADT) program was used to dock ligand molecules with receptor. A target-based docking method was used in this study to prepare the grid maps. The coordinate of the grid box was set at x = 24, y = 20 and z = 20, and the grid center was set at x = 8.638, y = 9.032 and z = 39.941. Docking simulation was executed with a number of modes set to 10 to get more accurate results. Ligand binding affinity was predicted as negative Gibbs free energy (∆G), which was calculated based on AutoDock Vina scoring function (Kcal/mol). The pose with the highest negative value is considered as the best conformation and selected for further analysis. Post-docking analysis was performed using LigPlot and PyMol. In the present study, mainly two types of interactions were analyzed, hydrogen bond and hydrophobic interaction.
Results and discussion
Curcumin and capsaicin, associated with 161 and 110 curated genes (supplementary data), were further subjected to functional enrichment analysis to obtain the cancer-associated targets. In cancer, curcumin and capsaicin have sensitized 34 and 35 differentially expressed genes (Table 1), which were used for network construction. The KEGG Mapper was used to map and identify the cancer-associated nodes of both curcumin and capsaicin molecules as shown in Fig. 1.
Network construction and analysis
Based on the STRING database evidence, the cancer-associated curcumin and capsaicin candidate interactive genes at high confidence level of 0.7 were considered to build the core protein–protein interaction using Cytoscape 3.2. The resultant network had 227 nodes/921 edges for curcumin and 223 nodes/954 edges for capsaicin, respectively, as nodes represent proteins and the edges indicate their relations. The network topology analysis identified ten hub proteins via the bottleneck method. The hub proteins in the interaction are a distinct scale-free network characterized by a power law distribution, which contains a small number of highly connected proteins known as hubs and a large number of poorly connected non-hub proteins. These hub proteins in the network strongly regulate cellular and biological functioning of the body than non-hub proteins. As a consequence of the limitations of the present approach, some human protein interactions were not comprehensive and found to be indistinct. As a result, highly connected network modules were selected for further analysis. Figure 2 represents the core interactome of curcumin targets in the cancer pathway and also the participants of angiogenic targets in the KEGG cancer signaling pathway. The resultant subnetwork of curcumin targets in cancer angiogenesis contains 112 nodes and 431 edges. Similarly, the subnetwork of capsaicin targets in cancer signaling events has 37 nodes and 85 edges. The participants of angiogenic targets of capsaicin in the KEGG cancer signaling pathway are shown in Fig. 3. The network topology analysis identified EP300, MAPK1, MAPK14, JUN, SRC, GSK3B, PTPN11 and STAT3 as potential angiogenic targets for both curcumin and capsaicin. Based on network topology analysis, MAPK1 possesses the highest degree of distribution and betweenness centrality and is thereby considered as a major driver of angiogenesis in cancer.
To evaluate the confidence of the protein interaction network, topological analysis was conducted using Network Analyzer v.3.3.1. Topological parameters such as betweenness centrality (BC), degree of distribution (D) and topological coefficient were computed to identify the bottleneck protein in the angiogenic interactome and the mainstream signaling pass through this node. This indicates the scale-free network of drug–protein interaction and possesses the property of modularity. The comparison of topological parameters of the giant network (core interactome) and subnetwork of both the molecules are presented in Table 2.
Gene Ontology (GO) analysis of MAPK1 first-degree neighboring genes identified in the angiogenic interactome of curcumin
The first-degree neighbors of hub nodes represented an extremely large percentage of the network driver nodes. To explore the functional characteristics of MAPK1, a network of first-degree neighbors of MAPK1 was created, subjected to gene ontology analysis and enriched according to the available GO annotations with gene ID and gene accessions in the PANTHER database. We retrieved all of the significant GO annotations including biological process (BP), molecular function (MF) and cellular component (CC), protein class and PANTHER pathway at a cutoff value of p ≤ 0.05 (Figs. 4, 5).
The GO enrichment analysis revealed that the first-degree neighbors of MAPK1 strongly interfere with the cellular process (GO 0009987) and metabolic processes (GO 0008152) and belong to the protein class of transferases (PC00220). In curcumin, these were mainly involved in the signaling process of CCKR signaling map (P06959), Ras pathway (P04393) and angiogenesis (P00005). Similarly, the GO enrichment analysis of first-degree nodes of MAPK1 in capsaicin molecule are mainly involved in cellular processes (GO 0009987) and metabolic processes (GO 0008152), which are under the protein class of transcription factor (PC00218) and transferases (PC00220) and are involved in the signaling event of CCKR signaling map (P06959), Ras pathway (P04393) and gonadotropin-releasing hormone receptor pathway (P06664) as well as angiogenesis (P00005) with less number of interacting partners. Hence, curcumin has been hypothetically predicted as a current target for MAPK1, which is further confirmed by molecular docking studies.
Molecular docking analysis of MAPK1 node with curcumin and capsaicin
Molecular docking analysis of curcumin, capsaicin and standard ulixertinib was performed using AutoDock Vina (ADT). After docking, the ligands were ranked according to their binding energy. ADT results were analyzed based on the interactions between receptor MAPK1 and ligand molecules (Table 3). The accuracies of the results were confirmed by considering the lowest binding free energy and the number of H bonds between the receptor and ligand. The binding energy of curcumin (− 7.6 kcal/mol) is very close to that of standard drug ulixertinib (− 8.1 kcal/mol), exhibiting H bonding (Table 3). Similarly, the binding energy of capsaicin (− 6.0 kcal/mol) against MAPK1 showed less inhibitory effect. Among the active pocket amino acid residues, Ile31, Gly37, Ala52, Lys54, Ile 56, Arg67, Leu107, Met108, Thr110, Asp111, Lys114, Val139, Leu156, Cys166 and Asp167 were actively involved in the formation of H bonds with MAPK1. The hydrogen and hydrophobic interaction profile of both curcumin and capsaicin with the receptor MAPK1 and their distances are shown in Fig. 6. Based on the results, it can be inferred that curcumin is a better inhibitor than capsaicin and may be considered for cancer treatment. In this context, curcumin should be further subjected to in vitro and in vivo validation to confirm these findings.
The topological analysis has led to the identification of MAPK1 as a top hub protein in regulating angiogenic signaling. This hypothesis was further examined by analyzing the binding mode of curcumin and capsaicin with MAPK1 by molecular docking studies. The docking of MAPK1 with curcumin (− 7.6 kcal/mol) has shown strong inhibition by exhibiting four hydrogen bonding compared to the standard ulixertinib (− 8.1 kcal/mol). Furthermore, functional enrichment analysis of first-degree neighbors of MAPK1 confirmed that the bottleneck proteins interact directly/indirectly with the non-hub proteins involved in regulating angiogenesis in cancer. Hence, the present study provides a potential biomarker known as bottleneck node as a therapeutic target for cancer treatment. However, further in vitro and in vivo experimental validations are needed to confirm these conclusions.
Assenov Y, Ramírez F, Schelhorn S-E, Lengauer T, Albrecht M (2008) Computing topological parameters of biological networks. Bioinformatics 24:282–284
Belenahalli Shekarappa S, Kandagalla S, Gollapalli P, BasavapattanaRudresh B, Hanumanthappa T, Hanumanthappa M (2017) Topology of protein–protein interaction network and edge reduction co-efficiency in VEGF signaling of breast cancer. Netw Model Anal Heal Inform Bioinform 6:17
Chin C-H, Chen S-H, Wu H-H, Ho C-W, Ko M-T, Lin C-Y (2014) cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 8:S11
Gaulton A, Bellis LJ, Bento AP, Chambers J, Davies M, Hersey A, Light Y, McGlinchey S, Michalovich D, Al-Lazikani B, Overington JP (2012) ChEMBL: a large-scale bioactivity database for drug discovery. Nucleic Acids Res 40:D1100–D1107
Hatcher H, Planalp R, Cho J, Torti FM, Torti SV (2008) Curcumin: from ancient medicine to current clinical trials. Cell Mol Life Sci 65:1631–1652
Huang D, Sherman BT, Tan Q, Collins JR, Alvord WG, Roayaei J, Stephens R, Baseler MW, Lane HC, Lempicki RA (2007) The DAVID gene functional classification tool: a novel biological module-centric algorithm to functionally analyze large gene lists. Genome Biol 8:R183
Kandagalla S, Sharath BS, Bharath BR, Hani U, Manjunatha H (2017) Molecular docking analysis of curcumin analogues against kinase domain of ALK5. Silico Pharmacol 5(1):15
Kanehisa M, Goto S (2000) KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 28:27–30
Kempaiah RK, Manjunatha H, Srinivasan K (2005) Protective effect of dietary capsaicin on induced oxidation of low-density lipoprotein in rats. Mol Cell Biochem 275:7–13
Kunnumakkara AB, Anand P, Aggarwal BB (2008) Curcumin inhibits proliferation, invasion, angiogenesis and metastasis of different cancers through interaction with multiple cell signaling proteins. Cancer Lett 269:199–225
Liu W, Wu A, Pellegrini M, Wang X (2015) Integrative analysis of human protein, function and disease networks. Sci Rep 5:14344
Manjunatha H, Srinivasan K (2006) Protective effect of dietary curcumin and capsaicin on induced oxidation of low-density lipoprotein, iron-induced hepatotoxicity and carrageenan-induced inflammation in experimental rats. FEBS J 273:4528–4537
Mi H, Poudel S, Muruganujan A, Casagrande JT, Thomas PD (2016) PANTHER version 10: expanded protein families and functions, and analysis tools. Nucleic Acids Res 44:D336–D342
Oyagbemi AA, Saba AB, Azeez OI (2010) Capsaicin: a novel chemopreventive molecule and its underlying molecular mechanisms of action. Indian J Cancer 47:53–58
Pyun B-J, Choi S, Lee Y, Kim T-W, Min J-K, Kim Y, Kim B-D, Kim J-H, Kim T-Y, Kim Y-M, Kwon Y-G (2008) Capsiate, a nonpungent capsaicin-like compound, inhibits angiogenesis and vascular permeability via a direct inhibition of Src kinase activity. Cancer Res 68:227–235
Raman K (2010) Construction and analysis of protein–protein interaction networks. Autom. Exp. 2:2
Ran J, Li H, Fu J, Liu L, Xing Y, Li X, Shen H, Chen Y, Jiang X, Li Y, Li H (2013) Construction and analysis of the protein-protein interaction network related to essential hypertension. BMC Syst Biol 7:32
Rose PW, Prlić A, Bi C, Bluhm WF, Christie CH, Dutta S, Green RK, Goodsell DS, Westbrook JD, Woo J, Young J, Zardecki C, Berman HM, Bourne PE, Burley SK (2015) The RCSB Protein Data Bank: views of structural biology for basic and applied research and education. Nucleic Acids Res 43:D345–D356
Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B, Ideker T (2003) Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 13:2498–2504
Sun PG, Gao L, Han S (2011) Prediction of human disease-related gene clusters by clustering analysis. Int. J. Biol. Sci. 7:61–73
Sung B, Prasad S, Yadav VR, Aggarwal BB (2012) Cancer cell signaling pathways targeted by spice-derived nutraceuticals. Nutr Cancer 64:173–197
Szklarczyk D, Franceschini A, Wyder S, Forslund K, Heller D, Huerta-Cepas J, Simonovic M, Roth A, Santos A, Tsafou KP, Kuhn M, Bork P, Jensen LJ, von Mering C (2015) STRING v10: protein–protein interaction networks, integrated over the tree of life. Nucleic Acids Res 43:D447–D452
Szklarczyk D, Santos A, von Mering C, Jensen LJ, Bork P, Kuhn M (2016) STITCH 5: augmenting protein-chemical interaction networks with tissue and affinity data. Nucleic Acids Res 44:D380–D384
Tang J, Aittokallio T (2014) Network pharmacology strategies toward multi-target anticancer therapies: from computational models to experimental design principles. Curr Pharm Des 20:23–36
Trott O, Olson AJ (2010) AutoDockVina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem 31:455–461
vanAalten DMF, Bywater R, Findlay JBC, Hendlich M, Hooft RWW, Vriend G (1996) PRODRG, a program for generating molecular topologies and unique molecular descriptors from coordinates of small molecules. J Comput Aided Mol Des 10:255–262
Wang Y, Xiao J, Suzek TO, Zhang J, Wang J, Bryant SH (2009) PubChem: a public information system for analyzing bioactivities of small molecules. Nucleic Acids Res 37:W623–W633
Xu X, Bi J, Ping L, Li P, Li F (2018) A network pharmacology approach to determine the synergetic mechanisms of herb couple for treating rheumatic arthritis. Drug Des Dev Ther 12:967–979
Zhang Y, Mao X, Guo Q, Lin N, Li S (2016) Network pharmacology-based approaches capture essence of Chinese herbal medicines. Chin Herb Med 8:107–116
The authors are thankful to the Registrar, Kuvempu University, Shankaraghatta-577 451 for providing facilities to complete this work.
Conflict of interest
The authors have declared that there is no conflict of interest.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Belenahalli Shekarappa, S., Kandagalla, S. & Hanumanthappa, M. A network pharmacology approach to investigate the pharmacological effect of curcumin and capsaicin targets in cancer angiogenesis by module-based PPI network analysis. J Proteins Proteom 10, 109–120 (2019). https://doi.org/10.1007/s42485-019-00012-y
- Network pharmacology
- Drug–target interaction
- Molecular docking