Exploring Molecular Targets and Mechanisms of Apigenin in the Treatment of Papillary Thyroid Carcinoma Based on Network Pharmacology and Molecular Docking Analysis

Abstract

To explore potential therapeutic targets and underlying mechanisms of apigenin (API) for papillary thyroid carcinoma (PTC) treatment. Methods: In the present study, network pharmacological analysis combined with molecular docking was used to identify API-related therapeutic targets for PTC treatment. Firstly, potential API-interacting and PTC-related targets were obtained from online databases. Furthermore, we constructed drug-target-disease interaction networks followed by functional enrichment analysis, expression patterns and prognostic values of predicted core targets by using bioinformatic tools. Moreover, we performed molecular docking to validate binding activity of API to core pharmacological targets. Results: A total of 110 API-related therapeutic targets were found to be engaged in the treatment of PTC. Moreover, TP53, HSP90AA1, AKT1, EGFR, SRC, VEGFA, ACTB, JUN and ESR1 were identified as core pharmacological targets of API for treating PTC. The enrichment analysis suggested that API played a comprehensive role in regulating the apoptotic process, modulating cell proliferation, and that multiple pathways were involved in API-related PTC treatment. Further results indicated that the PI3k-Akt/p53 signaling pathway was the key signaling pathway. Moreover, JUN was found mostly correlated with survival and molecular docking results demonstrating strong affinities between API and these core targets. Conclusion: Our findings systematically demonstrated API-associated pharmacological targets and mechanisms for treating PTC and demonstrated that API could function against PTC by inducing apoptosis, cell death and immunological reactions. Our research gives valuable insights and a theoretical basis for API-related PTC treatment.

Keywords

apigenin phytochemical papillary thyroid carcinoma network pharmacology enrichment analysis molecular docking

Introduction

Despite improving living standards, cancer is still one of the most serious life-threatening health problems and is the leading cause of death worldwide.¹ The number of new cancer cases in 2020 was 19.3 million, but the number is expected to be over 28.4 million by 2040,² which is an extreme burden on the whole society.³

Thyroid cancer (TC) is one of the most malignant endocrine system diseases.⁴ Papillary thyroid cancer (PTC) is the most prevalent type of thyroid carcinoma accounting for more than 85% of all diagnosed TC cases.⁵ Recently, with the wide application of high-resolution thyroid ultrasonography, the number of diagnosed PTC cases has increased significantly.⁶ Several studies indicated that genetic alterations, environmental factors, radiation exposure and family history of PTC may contribute to its development.⁷ In general, surgical treatment is still the first-line therapy for diagnosed PTC. Other clinical treatments include radiotherapy, chemotherapy, immunotherapy and combination therapy.⁸ However, excessive toxicity,⁹ radiation-related thyroid dysfunction¹⁰ and chemoresistance¹¹ of the existing treatment all remind us that we need to find more efficient and safer therapeutic approaches to PTC.

Phytochemicals can exert immense beneficial effects to our body.¹² Until now, researchers have identified nearly 10,000 phytochemicals, though many more are yet to be discovered.¹³ Polyphenols are secondary metabolites which account for the majority of phytochemicals.¹⁴ Recently, more and more attention has been paid to identify polyphenol's health beneficial properties.¹⁵ Apigenin (API; 4′,5,7-trihydroxyflavone), a bright yellow flavone, is abundant in many plants, such as celery, parsley and chamomile flowers.¹⁶ Recently, accumulating studies in vitro and in vivo have shown API's anti-inflammatory, anti-tumor, anti-apoptosis and cardioprotective effects.¹⁷ Sunanda Panda et al found that 0.78 mg/kg API significantly increased serum concentrations of thyroid hormone (T3 and T4) in alloxan-induced diabetic mice, and stimulated synthesis and secretion of thyroid hormone thus regulating thyroid function.¹⁸ Zhen Sun et al revealed that API played a chemotherapeutic role in human PTC BCPAP cells and significantly induced autophagy and ROS production in BCPAP cells.¹⁹ All these findings suggest that API may act as a promising drug in treating PTC. However, few studies have been carried out to identify the underlying mechanisms of API on PTC.

Network pharmacology is a newly emerging method that uses network analysis to identify a drug's biological ingredients and underlying therapeutic mechanisms.²⁰ Recently, in-silico techniques have become more and more powerful in drug research.^21–24 The application of network pharmacology and in-silico techniques gives us better insights into the effect of drugs in specific diseases.^25–28 In the present study, network pharmacological methods combined with bioinformatic tools were used to search for molecular interactions between API and PTC. Furthermore, an interaction network between API and PTC was constructed based on publicly available databases. Our study aimed to elucidate potential therapeutic targets and detailed therapeutic mechanisms of API for treating PTC and to provide a theoretical basis for PTC research.

Materials and Methods

Database Search for API-Interacting Targets

In the present study, API (Pubchem CID:5280443) was used as the ligand and was then retrieved from PubChem Database (https://pubchem.ncbi.nlm.nih.gov/).²⁹ The structure of API was downloaded in SDF format and was then converted to PDB format by using Open Babel software(http://ope-nbabel.org/wiki/Main_Page).³⁰ A 2D structure of API was presented by using ChemDraw software (Version 18.0).³¹ We searched for potential API-interacting targets in SwissTargetPrediction database (http://www.swisstargetprediction.ch/index.php), Traditional Chinese Medicine Systems Pha-rmacology Database (TCMSP, https://old.tcmspe.com/tcmsp.php), and Encyclopedia of Traditional Chinese Medicine database (ETCM, http://www.tc-mip.cn/ETCM/index.php).^25,32,33 The search results were pooled, and duplicates were removed.

Database Search for PTC-Related Targets

We searched for PTC-related targets in GeneCards database (https://www.genecards.org/), DisGeNET database (https://www.disgenet.org/) and Online Mendelian Inheritance in Man database (OMIM, https://omim.org/), using keywords ‘papillary thyroid carcinoma’ and ‘papillary thyroid cancer’ and species ‘Homo sapiens’.^34–36 The search results were pooled, and duplicates were removed.

Protein-Protein Interaction (PPI) Network Construction and Core Targets Identification

PPI networks are effective internet tools that are used for discovering biological interaction among proteins. Firstly, potential API-interacting targets and PTC-related targets were overlapped in a Venn diagram (http://bioinformatics.psb.ugent.be/webtools/Venn/) to find targets in common.³⁷ The intersected targets were considered potential therapeutic targets of API against PTC. STRING database (https://cn.string-db.org/), an online database, currently covers more than 24,000,000 proteins from 5000 organisms and enables researchers to construct functional associations for the proteins.^38,39 We then investigated functional association of these targets by using STRING database. The minimum interaction score was set as high confidence >0.7. Furthermore, cytoHubba plugin was used to predict core targets. Three different methods, namely MCC, MNC and Degree method, were applied in core target prediction. Cytoscape software (Version 3.8.0) was used to visualize the network analysis.⁴⁰

Functional Enrichment Analysis

In the present study, the Database for Annotation, Visualization and Integrated Discovery (D-AVID database, https://david.ncifcrf.gov/) was used to investigate the Gene Ontology(GO)-Biological Processes (BP) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis.⁴¹ Intersected targets were uploaded to DAVID database and the species was selected as Homo sapiens. KEGG analysis revealed crucial pathways involved in the predicted targets. A P value less than 0.05 was considered statistically significant. Additionally, Bioinformatic database (http://www.bioinfor-matics.com.cn/) was used to illustrate a target's bioinformatic annotation.

Construction of Gene-Pathway Network

To understand better the molecular mechanism involved in API-related PTC treatment, the gene-pathway network was performed after GO and KEGG enrichment analysis. The typical 20 analyzed BP and KEGG terms with contained genes that regulated these terms were visualized by using Cytoscape software.

Verification of Expression Patterns and Prognostic Values of Predicted Core Targets

The Human Protein Atlas (HPA database, https://www.proteinatlas.org/) is an open access database aimed at mapping all the human proteins in cells, tissues and organs.⁴² To increase further the credibility of our prediction, the HPA database was used to detect expression of predicted core targets in normal tissues and PTC tissues. Furthermore, the Gene Expression Profiling Interactive Analysis (GEPIA2) database was used to investigate potential interaction between disease-free survival (DFS) and predicted core targets.⁴³ In the present study, group cutoff was set as median and percentage of cutoff-high and cutoff-low was set as 50%. Furthermore, log-rank P < 0.05 was considered as statistically significant.

Molecular Docking

Molecular docking analysis was used to identify potential interaction between API and predicted targets. API(CID:5280443) was retrieved from PubChem database. Entry number of predicted proteins was obtained from Uniprot Database (https://www.uniprot.org/). 3D crystal structures of predicted macromolecules were obtained from Protein Data Bank database (PDB, https://www.pdbus.org/), whereas selection criterion was selected as Homo sapiens with refinement resolutions ranging from 1.0A to 3.0A. The collected original pdb format files were uploaded into Autodock and MGL tools for further analysis. Furthermore, ligand and macromolecules were prepared by using Pymol software (https://pymol.org/2/) including removal of water molecules, and charging addition and extraction of unwanted hetero-atoms. Additionally, Autodock (https://autodock.scripps.edu/) and Autodock tools 1.5.7 software (https://ccsb.scripps.edu/m-gltools/) were used to perform docking procedures between ligand and macromolecule using default parameters and the grid box was laid for the whole protein.⁴⁴ In the present study, the generated conformation with the best binding affinity was selected and visualized in Pymol and Discovery Studio 2019 software.^45,46 The ligand binds to proteins freely when the binding energy is less than 0. A lower binding energy score indicates a higher affinity of the ligand to target macromolecules.⁴⁷

Results and Discussion

Screening for Potential Therapeutic Targets of API for Treating PTC

The 2D structure of API is shown in Figure 1. A search for API-interacting targets was carried out in three databases, namely TCMSP, SwissTargetPrediction, and ETCM. The search yielded 80, 104, and 41 targets, respectively. These molecules were pooled together to obtain a total of 204 targets. A search for PTC-related targets was carried out in three online databases, namely GeneCards, DisGeNET, and OMIM. The search yielded 2918, 1348, and 634 targets, respectively. These targets were pooled together to get a total of 3826 targets. We then compared API-interacting targets with PTC-related targets and found 110 targets in common (Figure 2A). These were considered as potential therapeutic targets for API against PTC (Table 1). Gene network analysis of potential therapeutic targets of API for PTC treatment are shown in Figure 2B.

Figure 1.

Chemical structure of API.

Figure 2.

Network diagram of the overlapping targetsbetween API and PTC. (A)Gene network analysis of potential therapeutic targetsof API for treating PTC. The PPI network contains 96 nodes and 475 edges. (B)Network diagram of drug-disease target genes. The green V node represents API, the orange V node represents PTC, purple triangles represent unique API-related targetsand cyan nodes represent API-related therapeutic targets for treating PTC.

Table 1.

Putative Targets of API for Treating PTC.

No	Gene	Uniprot ID	No	Gene	Uniprot ID	No	Gene	Uniprot ID
1	ABCB1	P08183	38	CDK4	P11802	75	MMP13	P45452
2	ABCC1	P33527	39	CDK5R1	Q15078	76	MMP3	P08254
3	ABCG2	Q9UNQ0	40	CDK6	Q00534	77	MMP9	P14780
4	ACHE	P22303	41	CDKN2A	P42771	78	MPO	P05164
5	ACTB	P60709	42	CEBPB	P17676	79	NCOA1	Q15788
6	ADORA1	P30542	43	CFLAR	O15519	80	NEK2	P51955
7	ADORA2A	P29274	44	CFTR	P13569	81	NOX4	Q9NPH5
8	ADORA3	P0DMS8	45	CXCR1	P25024	82	NQO2	P16083
9	AHR	P35869	46	CYP19A1	P11511	83	P53	P04637
10	AHSA1	O95433	47	CYP1B1	Q16678	84	PARP1	P09874
11	AKR1B1	P15121	48	DAPK1	P53355	85	PFKFB3	Q16875
12	AKR1C3	P42330	49	DPP4	P27487	86	PIK3CG	P48736
13	AKT1	P31749	50	EGFR	P00533	87	PIK3R1	P27986
14	ALK	Q9UM73	51	ESR1	P03372	88	PIM1	P11309
15	ALOX12	P18054	52	ESR2	Q92731	89	PLG	P00747
16	ALOX15	P16050	53	ESRRA	P11474	90	PLK1	P53350
17	ALOX5	P09917	54	F2	P00734	91	PSME3	P61289
18	AMY1A	P0DUB6	55	FCER2	P06734	92	PTGS1	P23219
19	APC	P25054	56	FLT3	P36888	93	PTGS2	P35354
20	AR	P10275	57	GPER1	Q99527	94	PTK2	Q05397
21	AURKB	Q96GD4	58	GRK6	P43250	95	PTK2B	Q14289
22	AVPR2	P30518	59	GSK3B	P49841	96	PTPRS	Q13332
23	AXL	P30530	60	SHBG	P04278	97	SCN5A	Q14524
24	BAD	Q92934	61	ICAM1	P05362	98	HSP90AA1	P07900
25	BAX	Q07812	62	IGF1R	P08069	99	SLC5A5	Q92911
26	BCHE	P06276	63	IL13	P35225	100	SRC	P12931
27	BCL2	P10415	64	IL2	P60568	101	ST6GAL1	P15907
28	BCL2L1	Q07817	65	IL4	P05112	102	STK17B	O94768
29	CA12	O43570	66	INS	P01308	103	TERT	O14746
30	CA3	P07451	67	JAK1	P23458	104	TNF	P01375
31	CA4	P22748	68	JUN	P05412	105	TOP1	P11387
32	CA9	Q16790	69	KDR	P35968	106	TOP2A	P11388
33	CCNB1	P14635	70	MCL1	Q07820	107	TTR	P02766
34	CCND1	P24385	71	MDM2	Q00987	108	VEGFA	P15692
35	CD38	P28907	72	MET	P08581	109	XDH	P47989
36	CDK1	P06493	73	MMP1	P03956	110	MMP2	P08253
37	CDK2	P24941	74	MMP12	P39900

Drug-Disease-Interaction Network Construction

STRING database was used to construct a protein-protein interaction (PPI) network among 110 potential therapeutic targets. The resulting network was visualized by using Cytoscape software (Figure 3A). After removing outliers, the final PPI network contained 96 targets and 475 interactions, denoted by nodes and edges, respectively. Nodes with larger size and deeper color were more essential elements in the network. To identify the most essential API targets, we used three prediction methods, namely Maximal Clique Centrality (MCC), Maximum Neighborhood Component (MNC), and Degree. MCC, a newly proposed topological method, has a better performance in identifying essential proteins from complex PPI networks.⁴⁸ To ensure the accuracy of prediction, by overlapping prediction results from the three methods on a Venn diagram (Figure 3B), we obtained 9 targets in common. These were tumor protein p53(TP53), heat shock protein HSP90-alpha member 1(HSP90AA1), RAC-alpha serine/threonine-protein kinase (AKT1), epidermal growth factor receptor (EGFR), roto-oncogene tyrosine-protein kinase Src (SRC), vascular endothelial growth factor A (VEGFA), actin, cytoplasmic 1 (ACTB), transcription factor Jun (JUN) and estrogen receptor1 (ESR1) (Figure 3C). The degree of the top 9 targets is shown in Table 2 and they were considered core pharmacological targets and used for further analysis.

Figure 3.

Network analysis of core targetsof API for treating PTC. (A)Cytoscape analysis representing potential therapeutic targetsof API in the treatment of PTC. (B)Venn diagram of analysis results from CytoHubba(MCC, MNC, Degree). The overlapped parts representing core genes. (C) Network diagram of predicted core targets. Darker and redder represents higher degree and diamond nodes represent predicted core targets.

Table 2.

Degree of Top 9 Core Targets.

Rank	Gene Symbol	Annotation	Uniprot ID	Degree
1	TP53	Tumor Protein P53	P04637	44
2	HSP90AA1	Heat shock protein HSP 90-alpha member 1	P07900	38
3	AKT1	RAC-alpha serine/threonine-protein kinase	P31749	33
4	EGFR	Epidermal growth factor receptor	P00533	32
5	SRC	roto-oncogene tyrosine-protein kinase Src	P12931	31
6	VEGFA	Vascular endothelial growth factor A	P15692	30
7	ACTB	Actin, cytoplasmic 1	P60709	28
8	JUN	Transcription factor Jun	P05412	27
9	ESR1	Estrogen receptor1	P03372	25

Functional Enrichment Analysis

To understand better the functional mechanism of API in the treatment of PTC, DAVID database was used to perform enrichment analysis. We found a large number of enrichment terms including 371 BP terms and 112 KEGG terms that were significantly enriched. The top 20 BP and KEGG terms were visualized in a chord diagram (Figure 4A and 5A), bubble chart (Figure 4B and 5B) and histogram chart (Figure 4C and 5C) (P < 0.05). BP enrichment results suggested that the overlapping targets were involved in regulation of apoptotic process, positive regulation of phosphatidylinositol 3-kinase signaling, regulation of cell proliferation, and response to drug and protein phosphorylation. According to KEGG pathway analysis, the potential therapeutic targets were mostly enriched in hsa05200 (Pathways in cancer), hsa04151 (PI3K-AKT signaling pathway) and hsa04115 (p53 signaling pathway). The PI3K-AKT signaling pathway plays an important role in tumor apoptosis and the p53 signaling pathway is known to be involved in regulation of the apoptotic process, cell proliferation and DNA repair. Our KEGG enrichment analysis results suggested that API might exert positive effects on PTC by regulating the PI3K-AKT and p53 signaling pathways.

Figure 4.

Biological processes enrichment analysis of core targets. (A)The chord plot circle represents the top 20 enriched BP terms. The left grey part represents enriched targets and the colored ribbons represent enriched GO terms. (B)The bubble chart represents the top 20 enriched BP terms. The X-axis represents gene ratio, the Y-axis represents description and node sizes are represented as enriched gene counts. (C)The histogram represents the top 20 enriched BP terms. The X-axis represents enriched gene counts, the Y-axis represents description and -log10(pvalue) are represented as intensities of colors.

Figure 5.

KEGG enrichment analysis of core targets. (A)The chord plot circle represents the top 20 enriched KEGG terms. The left grey part represents enriched targets and the colored ribbons represent enriched KEGG terms. (B)The bubble chart represents the top 20 enriched KEGG terms. The X-axis represents gene ratio, the Y-axis represents description and node sizes are represented as enriched gene counts. (C)The histogram represents the top 20 enriched KEGG terms. The X-axis represents enriched gene counts, the Y-axis represents description and -log10(pvalue) are represented as intensities of colors.

Gene-Pathway Network Construction

The gene regulatory pathway network was constructed to identify the distribution of potential therapeutic targets and relevant pathways. The gene-pathway network is shown in Figure 6. In this network, predicted core targets and the related top 20 GO BP terms and KEGG signaling pathways were carried out with degree and visualized using Cytoscape software. Our results demonstrated that AKT1 had the most maximum degree and was considered as the core target gene. Other genes such as EGFR, SRC, VEGFA, TP53, and JUN also had larger degree, which further indicated that they might be core targets for the regulatory pathways of API against PTC. Enriched terms and corresponding contained genes are shown in Table 3 (BP) and Table 4 (KEGG).

Figure 6.

Gene-Pathway network of API for treating PTC. The orange V node represents PTC, the brown V node represents API, red circles represent core targets, purple parallelograms represent top 20 predicted KEGG pathway and pink rectangles represent top 20 predicted GO biological processes. Larger represents higher degree.

Table 3.

Enriched Biological Processes Relating to Core Targets.

Term	Biological process	Contained genes	P value
GO:0045893	positive regulation of transcription, DNA-templated	JUN, SRC, AKT1, ESR1, TP53, EGFR, ACTB	6.80E-08
GO:0042981	regulation of apoptotic process	HSP90AA1, AKT1, ESR1, TP53, ACTB	1.58E-06
GO:0033138	positive regulation of peptidyl-serine phosphorylation	HSP90AA1, AKT1, EGFR, VEGFA	5.62E-06
GO:0051897	positive regulation of protein kinase B signaling	HSP90AA1, SRC, EGFR, VEGFA	1.57E-05
GO:0071456	cellular response to hypoxia	SRC, AKT1, TP53, VEGFA	2.05E-05
GO:0043066	negative regulation of apoptotic process	SRC, AKT1, TP53, EGFR, VEGFA	3.53E-05
GO:0045944	positive regulation of transcription from RNA polymerase II promoter	JUN, AKT1, ESR1, TP53, EGFR, VEGFA	4.47E-05
GO:0001934	positive regulation of protein phosphorylation	HSP90AA1, AKT1, EGFR, VEGFA	6.78E-05
GO:0070374	positive regulation of ERK1 and ERK2 cascade	JUN, SRC, EGFR, VEGFA	8.78E-05
GO:0034614	cellular response to reactive oxygen species	JUN, AKT1, EGFR	9.99E-05
GO:0071276	cellular response to cadmium ion	JUN, AKT1, EGFR	1.23E-04
GO:0045429	positive regulation of nitric oxide biosynthetic process	HSP90AA1, AKT1, ESR1	1.76E-04
GO:0007173	epidermal growth factor receptor signaling pathway	SRC, AKT1, EGFR	1.76E-04
GO:0045737	positive regulation of cyclin-dependent protein serine/threonine kinase activity	SRC, AKT1, EGFR	1.76E-04
GO:0048146	positive regulation of fibroblast proliferation	JUN, ESR1, EGFR	2.38E-04
GO:0043065	positive regulation of apoptotic process	JUN, SRC, AKT1, TP53	2.57E-04
GO:0043154	negative regulation of cysteine-type endopeptidase activity involved in apoptotic process	SRC, AKT1, VEGFA	3.49E-04
GO:0001938	positive regulation of endothelial cell proliferation	JUN, AKT1, VEGFA	4.35E-04
GO:0043406	positive regulation of MAP kinase activity	SRC, EGFR, VEGFA	4.70E-04
GO:0050731	positive regulation of peptidyl-tyrosine phosphorylation	SRC, TP53, VEGFA	6.61E-04

Table 4.

Enriched Pathways Relating to Core Targets.

Term	Pathway	Contained genes	P value
hsa05418	Fluid shear stress and atherosclerosis	HSP90AA1, JUN, SRC, AKT1, TP53, ACTB, VEGFA;	5.99E-10
hsa05205	Proteoglycans in cancer	SRC, AKT1, ESR1, TP53, EGFR, ACTB, VEGFA;	6.30E-09
hsa05207	Chemical carcinogenesis - receptor activation	HSP90AA1, JUN, SRC, AKT1, ESR1, EGFR, VEGFA;	7.71E-09
hsa01522	Endocrine resistance	JUN, SRC, AKT1, ESR1, TP53, EGFR;	1.23E-08
hsa04915	Estrogen signaling pathway	HSP90AA1, JUN, SRC, AKT1, ESR1, EGFR;	6.94E-08
hsa04510	Focal adhesion	JUN, SRC, AKT1, EGFR, ACTB, VEGFA;	4.56E-07
hsa05131	Shigellosis	JUN, SRC, AKT1, TP53, EGFR, ACTB;	1.27E-06
hsa05200	Pathways in cancer	HSP90AA1, JUN, AKT1, ESR1, TP53, EGFR, VEGFA;	1.85E-06
hsa04919	Thyroid hormone signaling pathway	SRC, AKT1, ESR1, TP53, ACTB;	3.08E-06
hsa04926	Relaxin signaling pathway	JUN, SRC, AKT1, EGFR, VEGFA;	3.98E-06
hsa05219	Bladder cancer	SRC, TP53, EGFR, VEGFA;	6.49E-06
hsa05224	Breast cancer	JUN, AKT1, ESR1, TP53, EGFR;	6.70E-06
hsa05167	Kaposi sarcoma-associated herpesvirus infection	JUN, SRC, AKT1, TP53, VEGFA;	2.02E-05
hsa04015	Rap1 signaling pathway	SRC, AKT1, EGFR, ACTB, VEGFA;	2.76E-05
hsa05417	Lipid and atherosclerosis	HSP90AA1, JUN, SRC, AKT1, TP53;	3.03E-05
hsa05208	Chemical carcinogenesis - reactive oxygen species	JUN, SRC, AKT1, EGFR, VEGFA;	3.49E-05
hsa05163	Human cytomegalovirus infection	SRC, AKT1, TP53, EGFR, VEGFA;	3.62E-05
hsa05212	Pancreatic cancer	AKT1, TP53, EGFR, VEGFA;	4.21E-05
hsa01521	EGFR tyrosine kinase inhibitor resistance	SRC, AKT1, EGFR, VEGFA;	4.73E-05
hsa04012	ErbB signaling pathway	JUN, SRC, AKT1, EGFR;	5.89E-05

Expression Patterns and Survival Analysis of Predicted Core Targets

HPA and GEPIA databases were used to examine the expression pattern of API core targets in PTC patients and their relationship with survival rates. The group cutoff was set as median and the axis units were set as months. Significantly higher protein expression of ACTB, EGFR, HSP90AA1, AKT1, SRC, and VEGFA were observed in PTC tissues compared with normal tissues, whereas lower expression of JUN was observed in PTC tissues (Figure 7). Interestingly, lower expression of JUN was also correlated with lower survival rate of PTC patients (Figure 8F). Our findings suggested that JUN might serve as a potential biomarker for PTC treatment.

Figure 7.

Differential expressed analysis of predicted core targets from HPA database. Representative IHC results of (A)ACTB, (B)AKT1, (C)TP53, (D)EGFR, (E)ESR1, (F)VEGFA, (G)SRC, (H)JUN and (I)HSP90AA1 in normal and PTC tissues from HPA database.

Figure 8.

DFS analysis of predicted core genes from GEPIA2 database. Representative prognostic results of (A)ACTB, (B)AKT1, (C)EGFR, (D)ESR1, (E)HSP90AA1, (F)JUN, (G)SRC, (H)TP53 and (I)VEGFA in patients. The patients are divided into low(green) and high(red) group based on the median expression of predicted genes. Log-rank p < 0.05 is considered as statistically significant.

Molecular Docking Analysis

In this study, in silico analysis was used to identify biomolecule interaction between API and predicted proteins. Crystal structures of HSP90AA1(PDB ID: 7lt0), VEGFA (PDB ID: 4kzn), Akt1 (PDB ID: 7nh5), SRC (PDB ID: 1a09), Jun (PDB ID: 5t01), TP53 (PDB ID: 1aie), ESR1 (PDB ID: 1err), EGFR (PDB ID: 1m17) and ACTB (PDB ID: 6ict) were used as receptors for molecular docking.^49–57 Binding energy of API with predicted proteins ranged from −7.52 to −4.47 (Table 5) suggesting API had strong affinity with these proteins (Figure 9). Previous studies indicated that the affinity between curcumin and TP53 and VEGFA was −6.5 and −5 kcal/mol and the binding affinity between resveratrol and AKT1 was −6.2 kcal/mol.^58,59 Our study revealed that API showed better affinity with VEGFA (−5.27) and AKT1 (−6.83) in in silico studies. The co-crystalized ligand of the chosen macromolecule was re-docked using the same docking parameter as a test (Table 6). As for TP53, our results have shown that API possessed three conventional hydrogen bonds with ILE332, THR329 and ASN345, one carbon hydrogen bond with ILE332, and one pi-alkyl interaction ARG342; the residues LEU330, PHE341, GLU346, PHE338 and GLN331 were involved in van der Waals’ interactions with API. Additionally, API also showed strong binding affinity with AKT1 (−6.83) by forming one conventional hydrogen bond with THR211, SER205 and LYS268; three pi-pi stacked interactions with TRP80; one pi-anion interaction with ASP292; one pi-sigma interaction with LEU210; one pi-alkyl interaction with VAL270; the residues, including ASN 204, ALA212, ILE290, LEU264, and ASN53, took part in van der Waals’ interaction.

Figure 9.

Molecular view of best docking complexes between API and (A)TP53, (B)AKT1, (C)ACTB, (D)SRC, (E)HSP90, (F)VEGFA, (G)EGFR, (H)JUN, (I)ESR1. The API was hosted in the active site and was represented as green color. The macromolecule is represented as purple color and the binding sites are represented as cyan color. 2D intermolecular interactions of API with macromolecules are shown in the left.

Table 5.

Molecular Docking Score.

No	Target(PDB)	Binding sites with amino acid	Binding energy(kcal/mol)
1	HSP90AA1(7lt0)	ILE104, TRP162, ALA111, PHE138, GLY135, VAL136, LEU107, TYR139, GLY108, ILE26, GLN23, ALA21, PHE22, PHE170, LEU103	-6.73
2	VEGFA(4kzn)	GLY59, GLY58, CYS68, GLU67, LEU66, LYS107, GLU64, CYS60, ASP63, ASN62, CYS61	-5.27
3	AKT1(7nh5)	ASP292, ASN204, TRP80, SER205, VAL270, ASN53, LYS268, LEU264, ILE290, THR211, LEU210, ALA212	-6.83
4	SRC(1a09)	ASN224, SER225, LEU226, GLU149, TRP151, ILE146, GLN227, PHE215, LYS209, GLY213	-5.72
5	JUN(5t01)	ARG261, LYS258, MET253, HIS252, ALA265, ASN262	-4.47
6	TP53(1aie)	PHE341, GLU346, ILE332, PHE338, ASN345, ARG342, THR329, LEU330, GLN331	-5.48
7	ESR1(1err)	LEU525, GLY521, HIS524, LEU428, MET421, VAL422, PHE425, ILE424, ALA350, LEU349, GLU353, ARG394, LEU387, PHE404, LEU391, LEU384, MET388	-7.52
8	EGFR(1m17)	GLN958, MET963, ILE957, GLY959, ARG962, GLU961, ASN784, GLY786, ILE785, ASP960, ASP783, LYS782, PHE886	-5.63
9	ACTB(6ict)	TYR306, ALA331, ILE309, GLY308, ALA310, ASP311, PRO307, PRO333, GLU334, ARG335, PRO332, THR304, MET305,	-5.23

Table 6.

Molecular Docking Simulation Results.

Receptor (PDB ID)	Co-CrystallizedCompound Name	Co-CrystallizedCompound Structure	Co-Crystallizeddocking energy (kcal/mol)	API docking energy (kcal/mol)
7lt0	(1,3-dihydro-2H-isoindol-2-yl){3-[(3,4-dimethylphenyl)sulfanyl]-4-hydroxyphenyl}methanone		-10.4	-6.73
4kzn	TRIETHYLENE GLYCOL		-1.58	-5.27
7nh5	∼{N}-methyl-6-[4-[[4-[2-oxidanylidene-6-(propanoylamino)-3∼{H}-benzimidazol-1-yl]piperidin-1-yl]methyl]phenyl]-5-phenyl-pyridine-3-carboxamide		-6.9	-6.83
1a09	L-PEPTIDE LINKING		-5.32	-5.72
5t01	-	-	-	-4.47
1aie	-	-	-	-5.48
1err	RALOXIFENE		-11.8	-7.52
1m17	[6,7-BIS(2-METHOXY-ETHOXY)QUINAZOLINE-4-YL]-(3-ETHYNYLPHENYL)AMINE		-4.85	-5.63
6ict	S-ADENOSYL-L-HOMOCYSTEINE		-3.64	-5.23

Discussion

Thyroid cancer is the most prevalent endocrine system disease. PTC, the most common type of thyroid cancer, has gained attention in recent years due to its high incidence.⁶⁰ Though surgical removal is still the choice of treatment for PTC, its unexpected complications and side effects make it imperative to search for more reliable therapeutic strategies. The beneficial role that phytochemicals play in thyroid function has been fully discussed in recent years.⁶¹ As one of the most common type of phytochemicals, API has received much attention due to its low toxicity and potent biological activity.⁶² Though API is reported to be effective in the treatment of PTC, the central therapeutic targets and mechanisms of API against PTC still remain unclear.

In the present study, we created a network prediction of API-associated pharmacological targets and molecular mechanisms for treating PTC. To the best of our current knowledge, this is the first study revealing potential pharmacological targets and mechanisms of API against PTC by using network pharmacological methods. By using bioinformatic analysis, we obtained 110 targets which were related to PTC pathology and API treatment. Through the drug-disease interaction network, the top 9 targets with higher degree (TP53, HSP90AA1, AKT1, EGFR, SRC, VEGFA, ACTB, JUN and ESR1) were identified as core pharmacological targets related to PTC. Meanwhile, JUN was found to be closely related to the prognosis of PTC patients. According to our results, these core targets also showed better affinity with API in in silico studies, suggesting that API exerted its pharmacological effects by treating PTC via these targets. The enrichment analysis results demonstrated that regulation of the apoptotic process, positive regulation of phosphatidylinositol 3-kinase (PI3K) signaling, modulation of cell proliferation, and regulation of the PI3K-Akt/p53 signaling pathway might also be involved in the API treatment process. It appears that API could function against PTC by inducing apoptosis, cell death and immunological reactions.

TP53 displays the highest degree of protein-protein interaction among the identified core targets, which suggests that it could be the central target of API against PTC. This finding agrees with the existing observation that tp53 plays an essential role in regulating cell proliferation, apoptosis and DNA damage response. Previous studies have established a strong correlation between p53 expression and PTC progression. Elevated p53 expression was observed in diagnosed PTC patients.⁶³ Activation of the p53 signaling pathway was closely associated with the progression of PTC.⁶⁴ These findings are in agreement with our current results. A therapeutic approach targeting p53 and its associated signaling pathway has achieved preliminary success in preclinical studies.⁶⁵ Furthermore, API has been reported to induce apoptosis through a p53-independant signaling manner.⁶⁶ As an effective drug for treating PTC, API increases the stability of p53 at the same time and plays a vital role by targeting p53. A recent study reported that API induced G2/M phase cell cycle arrest and apoptosis in MCF-7 cells by regulating the p53 signaling pathway.⁶⁷ Our results imply that the regulation could be mediated by direct API-p53 interaction.

Aberrant activation of apoptosis-associated signaling pathway is central in PTC development. The PI3k-Akt signaling pathway is a well-known intracellular apoptotic pathway which is closely associated with the occurrence of PTC.⁶⁸ AKT1 is a key regulator in cell motility and the apoptotic process, the activation of which promotes metastasis in thyroid cancer.⁶⁹ Four out of 9 core targets identified in this study, namely VEGFA, AKT1, EGFR, and HSP90AA1, are in the PI3k-Akt signaling pathway. PI3K belongs to the lipid kinase family⁷⁰; activated PI3k induces phosphorylation of Akt and leads to the translocation of Akt from cytoplasm to the nucleus further regulating its downstream cellular process.⁷¹ Emerging evidence shows that inhibition of PI3k and its related pathway can be a novel target against cancer.⁷² Studies have shown that vitamin C dose-dependently inhibited the PI3k-Akt signaling pathway in BRAF thyroid cancer cells.⁷³ Thyroid hormone receptor beta (TRβ), a known PI3K inhibitor, exerts an anti-tumor effect in anaplastic thyroid cancer (ATC).⁷⁴ A study by Lu et al found that LncRNA HOTTIP deficiency inhibited cell proliferation, invasion and migration in PTC cells by targeting AKT1.⁷⁵ The pro-apoptosis effect of API in vitro was extensively verified. Furthermore, API has been shown to be effective in regulating the PI3K-AKT signaling pathway.⁷⁶ Yang et al reported that API could inhibit cell proliferation and induced apoptosis in hepatocellular carcinoma cells by inhibiting the PI3k/Akt/mTOR signaling pathway.⁷⁶ Thus, by regulating both p53 and the PI3K-AKT signaling pathway, our study indirectly demonstrates that API appears to be a promising and multitarget drug in the fight against PTC.

HSP90 is an upstream regulator of the AKT-related signaling pathway.⁷⁷ HSP90AA1, a stress inducible isoform of Hsp90, is reported to be involved in cell survival, cell circle control and apoptosis.⁷⁸ Under heat shock conditions, structural changes in cell membrane contribute to cell death. In PTC cell lines, expression of HSP90 is extremely high.⁷⁹ White et al found that HSP90 inhibitor could effectively inhibit migration and invasion of thyroid cancer stem cells.⁸⁰ Besides, API was reported to overcome the chemo-resistance of pancreatic cancer by targeting the mTOR-HSP90-mutp53-p62-NRF2-antioxidant response axis.⁸¹ The SRC oncogene has been proved to act as key regulator in tumor progression.⁸² Functionally, SRC binds to focal adhesion kinase (FAK) forming FAK/SRC complex and phosphorylates various downstream signaling molecules. SRC is also reported to phosphorylate the serine site of AKT and further affects its activity. By using an immunohistochemistry approach, a recent study found that compared with benign thyroid lesion, the percentage of positively SRC-stained cells was significantly high.⁸³ A recent study revealed that inhibition of the SRC-mediated PI3K pathway improved the overall survival of PTC patients with BRAF and RAS mutations.⁸⁴ Furthermore, treatment with API significantly attenuated SRC expression and exerted a chemopreventive activity against ultraviolet-induced skin inflammation.⁸⁵

EGFR is a transmembrane glycoprotein receptor belonging to the ErbB family of tyrosine kinase proteins. ErbB family members are associated with apoptosis and cell proliferation. Compared with benign thyroid tissues, a significantly increased mRNA level of EGFR was observed both in PTC tissues and cancer-adjacent tissues.⁸⁶ The BRAF V600E mutation is one of the most frequent genetic alterations in PTC patients. Targeting BRAF V600E appears to be a promising strategy for PTC patients exhibiting BRAF V600E mutations. Previous study demonstrated that treatment with EGFR inhibitor effectively enhanced the anticancer-effect of PLX4032 (a selective BRAF V600E inhibitor) in vitro.⁸⁷ What's more, Wang et al revealed that treatment with API could enhance the antitumor effects of cetuximab in nasopharyngeal carcinoma by inhibiting the EGFR signaling pathway.⁸⁸ These findings indicated that EGFR was involved in PTC development and API treatment.

The fact that PTC is more prevalent in females than in males implies possible involvement of estrogen in PTC progression. The estrogen receptor α gene ESR1 mediates the biological activity of estrogens and is reported to play a key role in PTC progression. Previous study indicated that increased level of ESR1 was observed in PTC tissues than in adjacent normal thyroid tissues.⁸⁹ Besides, lower overall survival is observed for PTC patients with higher ESR1 expression. Researchers have found that API could function as an effective antiestrogen in various type of cancer.⁹⁰ For instance, Nephew et al reported that API could inhibit the ESR1-dependant signaling pathway in breast cancer cell line.⁹¹ Furthermore, treatment with API significantly inhibited cervical cancer growth by reversing histamine-induced abnormal activation of ER signaling.⁹² Our findings suggest that API might exert antiestrogen effect in PTC tissues by targeting ESR1 and its-related signaling.

VEGFA is a member of the vascular endothelial growth factor (VEGF) family that regulates angiogenesis and vascular permeability, both of which are involved in tumor progression and metastasis. Angiogenesis is reported to be involved in PTC progression⁹³ and VEGFA has been reported to induce MMPs secretion, which can degrade the basal membrane.⁹⁴ Dysregulated permeability may lead to tumor expansion. Recent research suggested that inhibition of VEGF-mediated angiogenesis could be an effective way of PTC treatment. Salajegheh et al found that miR-126 significantly reduced VEGFA expression and inhibited cell proliferation in the PTC cancer cell line.⁹⁵ To some extent, API is a potent angiogenesis inhibitor. In non-small cell lung carcinoma, API significantly suppressed VEGF-A/VEGFR2 and PDGF-BB/PDGFβR signaling pathways and further reduced microvessel density.⁹⁶ Additionally, Ko et al found that in breast cancer cells, API suppressed secretion of VEGF and significantly overcame drug resistance via inhibition of the STAT3 signaling pathway.⁹⁷

c-JUN, one of the most well-studied transcription factors of the activator protein-1 (AP-1) family, is reported to be involved in a various of physiological processes including cell proliferation, immune activity and tumorigenesis. The role that c-JUN plays in the tumorigenesis of papillary carcinomas is well studied. A recent bioinformatics analysis indicated that c-JUN was closely relevant to PTC genesis.⁹⁸ Furthermore, Li et al found that increased expression of phosphorylated c-Jun NH2-terminal kinase (p-JNK) was highly associated with the presence of lymph node metastases.⁹⁹ Therefore, targeting the JNK/c-JUN signaling pathway may be a promising therapeutic intervention in PTC treatment. In the present study, JUN was identified as a core pharmacological target involved in API-related PTC treatment and was deduced as a prognosis marker for PTC patients. What's more, API is more likely to play a role by targeting JUN. Previous studies revealed that treatment with API (30 μM) significantly attenuated phosphorylation of JNK associated with MAPK signaling pathway in RAW264.7 cells.¹⁰⁰ Jeong et al found that pre-treatment with API and luteolin in HaCaT cells suppressed ultraviolet irradiation-induced expression of c-JUN.¹⁰¹ Thus, API targeting JUN may provide totally fresh insights and improve prognosis in the treatment of PTC.

In conclusion, API is expected to be a promising and effective drug for treating PTC. Despite all these novelties, there are several limitations to this study. On one hand, network pharmacological methods are based on the existing network databases, and in-depth mechanisms and relevant results still require constant updating. On the other hand, further pharmacological verifications are still required to verify our results and we will continue to investigate API's role in the treatment of PTC. Finally, much remains to be done to fully understand the pathogenesis of PTC and transfer our findings to clinical practice.

Conclusions

Taken together, in this study we systematically elucidated possible targets and potential therapeutic pathways of API in the treatment of PTC by using network pharmacology, bioinformatics analysis and molecular docking. Based on the network pharmacological results, 9 effective pharmacological targets (TP53, HSP90AA1, AKT1, EGFR, SRC, VEGFA, ACTB, JUN and ESR1) were identified as therapeutic targets of API against PTC, and molecular docking results also suggested that API could bind with these targets stably, further providing a theoretical basis to investigate the pathogenesis of PTC. Meanwhile, JUN served as a new prognosis of PTC. The functional enrichment results strongly suggest that therapeutic mechanisms of API against PTC might be associated with regulation of the apoptotic process, mediating cell proliferation, immunomodulation and regulating the PI3k-Akt/p53 signaling pathway. Our study provides valuable insights and a theoretical basis for further investigation of API for treating PTC.

Footnotes

Acknowledgements

The authors thank Dr Zhiyun Chen (Department of Microbiology and Immunology, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, USA) for the English editing of the manuscript.

Author's Contribution

All authors contributed to the revision of the study. Dongyu Li carried out the study and contributed to data collection and analysis. Lei Wang and Yuchen Jing contributed to manuscript writing. Bo Jiang, Lei Zhao and Yuxi Miao contributed to data analysis. Shijie Xin and Chunlin Ge contributed to manuscript corrections. All authors approved final manuscript to be published.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Availability of Data and Materials

The data supporting the findings of this study are available within the article.

ORCID iD

Dongyu Li

References

Torre

Siegel

Ward

Jemal

. Global cancer incidence and mortality rates and trends--an update. Cancer Epidemiol Biomarkers Prev: Publ Am Assoc Cancer Res Cosponsored Am Soc Preventive Oncol. 2016;25(1):16-27.

Sung

Ferlay

Siegel

, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2021;71(3):209-249.

Lortet-Tieulent

Georges

Bray

Vaccarella

. Profiling global cancer incidence and mortality by socioeconomic development. Int J Cancer 2020;147(11):3029-3036.

Cabanillas

McFadden

Durante

. Thyroid cancer. Lancet (London, England). 2016;388(10061):2783-2795.

Shin

Hwang

Choi

. Antitumor effect of various phytochemicals on diverse types of thyroid cancers. Nutrients. 2019;11(1):125.

Yuan

Mirshahidi

, et al. Thyroid carcinoma: Phenotypic features, underlying biology and potential relevance for targeting therapy. Int J Mol Sci. 2021;22(4):1950.

Abdullah

Junit

, et al. Papillary thyroid cancer: genetic alterations and molecular biomarker investigations. Int J Med Sci 2019;16(3):450-460.

Zheng

Guo

. AHNAK2 Is a novel prognostic marker and correlates with immune infiltration in papillary thyroid cancer: evidence from integrated analysis. Int Immunopharmacol 2021;90:107185.

Laha

Nilubol

Boufraqech

. New therapies for advanced thyroid cancer. Front Endocrinol (Lausanne) 2020;11:82.

10.

Albi

Cataldi

Lazzarini

, et al. Radiation and thyroid cancer. Int J Mol Sci 2017;18(5):911.

11.

Bulotta

Capriglione

Celano

, et al. Phytochemicals in thyroid cancer: analysis of the preclinical studies. Endocrine. 2021;73(1):8-15.

12.

Pagliaro

Santolamazza

Simonelli

Rubattu

. Phytochemical compounds and protection from cardiovascular diseases: a state of the art. BioMed Res Int 2015;2015:918069.

13.

Zhang

Gan

, et al. Antioxidant phytochemicals for the prevention and treatment of chronic diseases. Molecules. 2015;20(12):21138-21156.

14.

Bowtell

Kelly

. Fruit-derived polyphenol supplementation for athlete recovery and performance. Sports Med (Auckland, N.Z.). 2019;49(Suppl 1):3-23.

15.

Ganesan

. Polyphenol-rich lentils and their health promoting effects. Int J Mol Sci. 2017;18(11):2390.

16.

Tang

Chen

Huang

. Pharmacokinetic properties and drug interactions of apigenin, a natural flavone. Expert Opin Drug Metab Toxicol 2017;13(3):323-330.

17.

Salehi

Venditti

Sharifi-Rad

, et al. The therapeutic potential of apigenin. Int J Mol Sci. 2019;20(6):1305.

18.

Panda

Kar

. Apigenin (4’,5,7-trihydroxyflavone) regulates hyperglycaemia, thyroid dysfunction and lipid peroxidation in alloxan-induced diabetic mice. J Pharm Pharmacol 2007;59(11):1543-1548.

19.

Zhang

Cheng

Gao

, et al. Apigenin induces autophagic cell death in human papillary thyroid carcinoma BCPAP cells. Food Funct. 2015;6(11):3464-3472.

20.

Berger

Iyengar

. Network analyses in systems pharmacology. Bioinformatics (Oxford. England). 2009;25(19):2466-2472.

21.

Singh

Bhardwaj

Das

Purohit

. Identification of 11β-HSD1 inhibitors through enhanced sampling methods. Chem Commun (Cambridge. England). 2022;58(32):5005-5008.

22.

Bhardwaj

Oakley

Purohit

. Mechanistic behavior and subtle key events during DNA clamp opening and closing in T4 bacteriophage. Int J Biol Macromol 2022;208:11-19.

23.

Kumar

Bhardwaj

Singh

Das

Purohit

. Identification of acridinedione scaffolds as potential inhibitor of DENV-2 C protein: an in silico strategy to combat dengue. J Cell Biochem 2022;123(5):935-946.

24.

Bhardwaj

Purohit

. A lesson for the maestro of the replication fork: targeting the protein-binding interface of proliferating cell nuclear antigen for anticancer therapy. J Cell Biochem 2022;123(6):1091-1102.

25.

Zhang

Zhu

Bai

Ning

. Network pharmacology databases for Traditional Chinese Medicine: review and assessment. Front Pharmacol. 2019;10:123.

26.

Rajendran

Purohit

Sethumadhavan

. In silico investigation of molecular mechanism of laminopathy caused by a point mutation (R482W) in lamin A/C protein. Amino Acids. 2012;43(2):603-615.

27.

Rajendran

Sethumadhavan

. Drug resistance mechanism of PncA in Mycobacterium tuberculosis. J Biomol Struct Dyn 2014;32(2):209-221.

28.

Singh

Bhardwaj

Purohit

. Computational targeting of allosteric site of MEK1 by quinoline-based molecules. Cell Biochem Funct 2022;40(5):481-490.

29.

Kim

Chen

Cheng

, et al. Pubchem in 2021: new data content and improved web interfaces. Nucleic Acids Res 2021;49(D1):D1388-d1395.

30.

O’Boyle

Morley

Hutchison

. Pybel: a python wrapper for the OpenBabel cheminformatics toolkit. Chem Cent J 2008;2:5.

31.

Evans

. History of the harvard ChemDraw project. Angew Chem (Int ed. Engl). 2014;53(42):11140-11145.

32.

Gfeller

Grosdidier

Wirth

, et al. Swisstargetprediction: a web server for target prediction of bioactive small molecules. Nucleic Acids Res 2014;42(Web Server issue):W32-W38.

33.

Zhang

Liu

, et al. ETCM: an encyclopaedia of traditional Chinese medicine. Nucleic Acids Res 2019;47(D1):D976-d982.

34.

Stelzer

Rosen

Plaschkes

, et al. The GeneCards Suite: from gene data mining to disease genome sequence analyses. Current Protoc Bioinf. 2016;54:1.30.31-31.30.33.

35.

Piñero

Ramírez-Anguita

Saüch-Pitarch

, et al. The DisGeNET knowledge platform for disease genomics: 2019 update. Nucleic Acids Res 2020;48(D1):D845-d855.

36.

Amberger

Hamosh

. Searching online Mendelian inheritance in man (OMIM): a knowledgebase of human genes and genetic phenotypes. Current Protoc in Bioinf. 2017;58:1.2.1-1.2.12.

37.

Jia

Wang

. Venn diagrams in bioinformatics. Brief Bioinform 2021;22(5):bbab108.

38.

Priyamvada

Debroy

Anbarasu

Ramaiah

. A comprehensive review on genomics, systems biology and structural biology approaches for combating antimicrobial resistance in ESKAPE pathogens: computational tools and recent advancements. World J Microbiol Biotechnol 2022;38(9):153.

39.

Basu

Veeraraghavan

Ramaiah

Anbarasu

. Novel cyclohexanone compound as a potential ligand against SARS-CoV-2 main-protease. Microb Pathog 2020;149:104546.

40.

Shannon

Markiel

Ozier

, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003;13(11):2498-2504.

41.

Dennis

Jr. Sherman

Hosack

, et al. DAVID: database for annotation, visualization, and integrated discovery. Genome Biol 2003;4(5):P3.

42.

Colwill

Gräslund

. A roadmap to generate renewable protein binders to the human proteome. Nat Methods 2011;8(7):551-558.

43.

Tang

Kang

Chen

Zhang

. GEPIA2: An enhanced web server for large-scale expression profiling and interactive analysis. Nucleic Acids Res 2019;47(W1):W556-w560.

44.

Liang

, et al. Network pharmacology and bioinformatics analyses identify intersection genes of niacin and COVID-19 as potential therapeutic targets. Brief Bioinform 2021;22(2):1279-1290.

45.

Seeliger

de Groot

. Ligand docking and binding site analysis with PyMOL and autodock/vina. J Comput-Aided Mol Des 2010;24(5):417-422.

46.

Zhu

Zhong

Peng

. Study on the mechanism of baimai ointment in the treatment of osteoarthritis based on network pharmacology and molecular docking with experimental verification. Front Genet 2021;12:750681.

47.

Hopkins

Keserü

Leeson

Rees

Reynolds

. The role of ligand efficiency metrics in drug discovery. Nat Rev Drug Discov 2014;13(2):105-121.

48.

Chin

Chen

, et al. Cytohubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 2014;8(Suppl 4):S11.

49.

Mishra

Khandelwal

Banerjee

, et al. Selective inhibition of the Hsp90α isoform. Angew Chem (Int ed. Engl). 2021;60(19):10547-10551.

50.

Quambusch

Depta

Landel

, et al. Cellular model system to dissect the isoform-selectivity of Akt inhibitors. Nat Commun. 2021;12(1):5297.

51.

Wang

Liu

, et al. Resveratrol: multi-targets mechanism on neurodegenerative diseases based on network pharmacology. Front Pharmacol. 2020;11:694.

52.

Hong

Wang

Horton

, et al. Methyl-dependent and spatial-specific DNA recognition by the orthologous transcription factors human AP-1 and Epstein-Barr virus Zta. Nucleic Acids Res 2017;45(5):2503-2515.

53.

Charifson

Shewchuk

Rocque

, et al. Peptide ligands of pp60(c-src) SH2 domains: a thermodynamic and structural study. Biochemistry. 1997;36(21):6283-6293.

54.

Mittl

Chène

Grütter

. Crystallization and structure solution of p53 (residues 326–356) by molecular replacement using an NMR model as template. Acta Crystallogr D Biol Crystallogr 1998;54(Pt 1):86-89.

55.

Brzozowski

Pike

Dauter

, et al. Molecular basis of agonism and antagonism in the oestrogen receptor. Nature. 1997;389(6652):753-758.

56.

Guo

Liao

Kwiatkowski

, et al. Structural insights into SETD3-mediated histidine methylation on β-actin. eLife. 2019;8:e43676.

57.

Stamos

Sliwkowski

Eigenbrot

. Structure of the epidermal growth factor receptor kinase domain alone and in complex with a 4-anilinoquinazoline inhibitor. J Biol Chem 2002;277(48):46265-46272.

58.

Kang

Zhong

, et al. A network pharmacology and molecular docking strategy to explore potential targets and mechanisms underlying the effect of curcumin on osteonecrosis of the femoral head in systemic lupus erythematosus. BioMed Res Int 2021;2021:5538643.

59.

Liu

Dong

Wang

Wandong

. Resveratrol ameliorates lipopolysaccharide-induced sudden sensorineural hearing loss in in vitro model through multitarget antiapoptotic mechanism based on network pharmacology and molecular docking. Evidence-based Complementary Alternative Med: eCAM. 2022;2022:6404588.

60.

McLeod

DSA

Zhang

Durante

Cooper

. Contemporary debates in adult papillary thyroid cancer management. Endocr Rev 2019;40(6):1481-1499.

61.

Paunkov

Chartoumpekis

Ziros

, et al. Impact of antioxidant natural compounds on the thyroid gland and implication of the Keap1/Nrf2 signaling pathway. Curr Pharm Des 2019;25(16):1828-1846.

62.

Sharma

Ghani

Sak

, et al. Probing into therapeutic anti-cancer potential of apigenin: recent trends and future directions. Recent Pat Inflammation Allergy Drug Discovery 2019;13(2):124-133.

63.

Radu

Mogoantă

Busuioc

Stănescu

Grosu

. Histological and immunohistochemical aspects of papillary thyroid cancer. Rom J Morphol Embryol= Revue roumaine de morphologie et embryologie. 2015;56(2 Suppl):789-795.

64.

Shih

Davis

Lin

Davis

. Resveratrol induces apoptosis in thyroid cancer cell lines via a MAPK- and p53-dependent mechanism. J Clin Endocrinol Metab 2002;87(3):1223-1232.

65.

Lan

Zhang

Wang

, et al. Genome-wide analysis of long noncoding RNA expression profile in papillary thyroid carcinoma. Gene. 2015;569(1):109-117.

66.

Meng

Zhu

, et al. Apigenin inhibits renal cell carcinoma cell proliferation. Oncotarget. 2017;8(12):19834-19842.

67.

Shendge

Chaudhuri

Basu

Mandal

. A natural flavonoid, apigenin isolated from Clerodendrum viscosum leaves, induces G2/M phase cell cycle arrest and apoptosis in MCF-7 cells through the regulation of p53 and caspase-cascade pathway. Clin Transl Oncol: Off Publ Fed Span Oncol Soc National Cancer Inst Mexico. 2021;23(4):718-730.

68.

Amjad

Asnaashari

Sokouti

. The role of PI3K signaling pathway and its associated genes in papillary thyroid cancer. J Egypt Natl Canc Inst 2021;33(1):11.

69.

Saji

Narahara

McCarty

, et al. Akt1 deficiency delays tumor progression, vascular invasion, and distant metastasis in a murine model of thyroid cancer. Oncogene. 2011;30(42):4307-4315.

70.

Liu

Chen

Liu

, et al. PI3K/AKT Pathway as a key link modulates the multidrug resistance of cancers. Cell Death Dis 2020;11(9):797.

71.

Yang

Jiang

Hou

. Emerging role of PI3K/AKT in tumor-related epigenetic regulation. Semin Cancer Biol 2019;59:112-124.

72.

LoRusso

. Inhibition of the PI3K/AKT/mTOR pathway in solid tumors. J Clin Oncol: Off J Am Soc Clin Oncol. 2016;34(31):3803-3815.

73.

Shen

Yang

, et al. Vitamin C kills thyroid cancer cells through ROS-dependent inhibition of MAPK/ERK and PI3K/AKT pathways via distinct mechanisms. Theranostics. 2019;9(15):4461-4473.

74.

Bolf

Gillis

Davidson

, et al. Thyroid hormone receptor beta induces a tumor-suppressive program in anaplastic thyroid cancer. Mol Cancer Res: MCR. 2020;18(10):1443-1452.

75.

Yuan

Liu

Fan

, et al. LncRNA HOTTIP promotes papillary thyroid carcinoma cell proliferation, invasion and migration by regulating miR-637. Int J Biochem Cell Biol 2018;98:1-9.

76.

Yang

Wang

. Inhibition of PI3K/akt/mTOR pathway by apigenin induces apoptosis and autophagy in hepatocellular carcinoma cells. Biomed Pharmacother = Biomedecine & Pharmacotherapie. 2018;103:699-707.

77.

Abeyrathna

. The critical role of akt in cardiovascular function. Vasc Pharmacol 2015;74:38-48.

78.

Wang

Zou

, et al. Interaction of Hsp90AA1 with phospholipids stabilizes membranes under stress conditions. Biochimica et biophysica acta. Biomembranes. 2019;1861(2):457-465.

79.

Choi

Jeong

Jeon

Ahn

. HIF-1α and HSP90: target molecules selected from a tumorigenic papillary thyroid carcinoma cell line. Cancer Sci 2012;103(3):464-471.

80.

White

Subramanian

Zhu

, et al. Novel HSP90 inhibitors effectively target functions of thyroid cancer stem cell preventing migration and invasion. Surgery. 2016;159(1):142-151.

81.

Montani

Cecere

Granato

et al.

. Mutant p53, stabilized by its interplay with HSP90, activates a positive feed-back loop between NRF2 and p62 that induces chemo-resistance to apigenin in pancreatic cancer cells. Cancers (Basel). 2019;11(5):703.

82.

Simatou

Simatos

Goulielmaki

, et al. Historical retrospective of the SRC oncogene and new perspectives (Review). Mol Clin Oncol 2020;13(4):21.

83.

Michailidi

Giaginis

Stolakis

, et al. Evaluation of FAK and src expression in human benign and malignant thyroid lesions. Pathol Oncol Res: POR. 2010;16(4):497-507.

84.

Beadnell

Nassar

Rose

, et al. Src-mediated regulation of the PI3K pathway in advanced papillary and anaplastic thyroid cancer. Oncogenesis. 2018;7(2):23.

85.

Byun

Park

Lee

, et al. Src kinase is a direct target of apigenin against UVB-induced skin inflammation. Carcinogenesis. 2013;34(2):397-405.

86.

Lam

Lau

Gopalan

Luk

. Quantitative analysis of the expression of TGF-alpha and EGFR in papillary thyroid carcinoma: clinicopathological relevance. Pathology. 2011;43(1):40-47.

87.

Jia

Zhang

, et al. EGFR Inhibition enhances the antitumor efficacy of a selective BRAF V600E inhibitor in thyroid cancer cell lines. Oncol Lett 2018;15(5):6763-6769.

88.

Liu

Zhong

Wang

. Apigenin enhances the antitumor effects of cetuximab in nasopharyngeal carcinoma by inhibiting EGFR signaling. Biomed Pharmacother = Biomedecine & Pharmacotherapie. 2018;102:681-688.

89.

Fan

Liu

van Hasselt

, et al. Estrogen receptor α induces prosurvival autophagy in papillary thyroid cancer via stimulating reactive oxygen species and extracellular signal regulated kinases. J Clin Endocrinol Metab 2015;100(4):E561-E571.

90.

Yao

Fan

Han

Sun

Che

. Apigenin acts as a partial agonist action at estrogen receptors in vivo. Eur J Pharmacol 2021;906:174175.

91.

Long

Fan

Bigsby

Nephew

. Apigenin inhibits antiestrogen-resistant breast cancer cell growth through estrogen receptor-alpha-dependent and estrogen receptor-alpha-independent mechanisms. Mol Cancer Ther 2008;7(7):2096-2108.

92.

Zhang

Fan

, et al. Apigenin inhibits histamine-induced cervical cancer tumor growth by regulating estrogen receptor expression. Molecules. 2020;25(8):1960.

93.

Peng

Chen

Zhang

, et al. VEGF Trapon inhibits tumor growth in papillary thyroid carcinoma. Eur Rev Med Pharmacol Sci 2015;19(2):235-240.

94.

Claesson-Welsh

Welsh

. VEGFA And tumour angiogenesis. J Intern Med 2013;273(2):114-127.

95.

Salajegheh

Vosgha

Rahman

, et al. Interactive role of miR-126 on VEGF-A and progression of papillary and undifferentiated thyroid carcinoma. Hum Pathol 2016;51:75-85.

96.

Zeng

Chen

, et al. Apigenin suppresses tumor angiogenesis and growth via inhibiting HIF-1α expression in non-small cell lung carcinoma. Chem-Biol Interact 2022;361:109966.

97.

Seo

Choi

, et al. Apigenin overcomes drug resistance by blocking the signal transducer and activator of transcription 3 signaling in breast cancer cells. Oncol Rep 2017;38(2):715-724.

98.

Chen

Liu

, et al. Special role of JUN in papillary thyroid carcinoma based on bioinformatics analysis. World J Surg Oncol 2017;15(1):119.

99.

Wang

Chao

Zhen

, et al. Phosphorylated c-Jun NH2-terminal kinase is overexpressed in human papillary thyroid carcinomas and associates with lymph node metastasis. Cancer Lett 2010;293(2):175-180.

100.

Park

Min

, et al. Effects of apigenin on RBL-2H3, RAW264.7, and HaCaT cells: anti-allergic, anti-inflammatory, and skin-protective activities. Int J Mol Sci. 2020;21(13):4620.

101.

Lin

Wang

, et al. Apigenin attenuates dopamine-induced apoptosis in melanocytes via oxidative stress-related p38, c-jun NH2-terminal kinase and akt signaling. J Dermatol Sci 2011;63(1):10-16.