To Explore the Inhibitory Mechanism of Quercetin in Thyroid Papillary Carcinoma through Network Pharmacology and Experiments

Quercetin, a flavonoid with anti-inflammatory and anticancer properties, is expected to be an innovative anticancer therapeutic agent for papillary thyroid carcinoma (PTC). However, the downstream signaling pathways that mediate quercetin-dependent anticancer properties remain to be deciphered. Herein, potential targets of quercetin were screened with several bioinformatic avenues including PharmMapper, Gene Expression Omnibus (GEO) database, protein–protein interaction (PPI) network, and molecular docking. Besides, western blot, CCK-8 transwell analysis of migration and invasion, flow cytometric analysis, and colony formation assays were performed to investigate the underlying mechanism. We found four core nodes (MMP9, JUN, SPP1, and HMOX1) by constructing a PPI network with 23 common targets. Through functional enrichment analysis, we confirmed that the above four target genes are enriched in the TNF, PI3K-AKT, and NF-κB signaling pathways, which are involved in the inflammatory microenvironment and inhibit the development and progression of tumors. Furthermore, molecular docking results demonstrated that quercetin shows strong binding efficiency with the proteins encoded by these 4 key proteins. Finally, quercetin displayed strong antitumor efficacy in PTC cell lines. In this research, we demonstrated the application of network pharmacology in evaluating the mechanisms of action and molecular targets of quercetin, which regulates a variety of proteins and signaling pathways in PTC. These data might explain the mechanism underlying the anticancer effects of quercetin in PTC.


Introduction
Thyroid cancer, the most common malignant tumor of the endocrine system, is one of the fastest growing tumors in the world [1] and is increasing 3.4% annually [2]. Papillary thyroid cancer (PTC) is the most common type of thyroid cancer, and most patients have a good prognosis after surgery, thyroid-stimulating hormone suppression, and/or radioactive iodine therapy [3]. However, 5% of patients finally get distant metastasis [4] which means worse prognosis. The reasons are still unclear, and earlier detection of PTC is urgently needed [5]. The main reason for the progression of these patients is probably the dedifferentiation of the tumor cells. Due to dedifferentiation of tumor cells, patients have reduced expression and abnormal function of thyroid-stimulating hormone receptor (TSHR) and natrium iodide symporter (NIS), resulting in insensitivity to 131I therapy and thus becoming iodine-refractory differentiated thyroid cancer with poor prognosis. Recent studies have suggested the relationship between the immune microenvironment and tumor aggressiveness [6].
Quercetin is a flavonoid that is found in various fruits and vegetables. It has antiviral, antioxidant, and antiallergic properties and a variety of pharmacological effects [7]. Moreover, quercetin is well known for its anticarcinogenic potential; the mechanism is involved in intercalating DNA in cancer cell lines [8] or binding to cellular receptors [9]. Previous studies have shown the cancer prevention and treatment function of quercetin in colon cancer [10] and cervical cancer [11].
Network pharmacology is a method for drug design that uncovers complex relationships between multiple components and targets, and it contains network analysis, systems biology, redundancy, connectivity, and pleiotropy [12]. Various studies have demonstrated the power of network analysis in understanding biological systems. Network pharmacology enables effective elucidating of previously unexplored natural products and provides systematic methods to extend the space of drug treatment in various diseases [13].
As quercetin exerts multiple pharmacological effects and multiple targets, therefore, we try to find a method by which we can systematically elucidate the role of quercetin. Protein-protein interaction (PPI) applies to the process by which protein molecules constitute a protein complex by noncovalent bonds. The interactions between drugs and specific nodes or modules can be elucidated to identify potential mechanisms using PPI network [14]. Besides, we performed molecular docking for determining protein-ligand interactions. In the study, we sought to verify the role of quercetin on PTC and the associated signaling pathway through network pharmacology and then verified the proteins in the pathway to provide rationale for anti-PTC therapy. The roadmap of this study is shown in Figure 1(a).

Results
2.1. DEGs in PTC. 105 samples were downloaded from the GEO database, which were categorized as two groups: the tumor group including 60 PTC samples and the normal group including 45 normal thyroid samples. Compared to the normal group, 904 DEGs were obtained in the cancer group, containing 481 downregulated genes and 423 upregulated ones. The volcano plot showed the DEGs identified between normal group and tumor group samples (Figure 1(b)). The heat map showed the expression of the DEGs (Figure 1(c)). The 904 DEGs were cross-fertilized with the quercetin targets analyzed by the database for subsequent construction of the protein network.
2.2. Potential Targets of Quercetin in PTC. 144 targets of quercetin were obtained by utilizing the online database, and 23 common targets, which are relevant to both quercetin and PTC, were filtered out by Venn diagram (Figure 2(b)). The 23 targets were used for the establishment of PPI network.
To clarify the connections among the 23 target proteins, we constructed a PPI network (Figure 2(c)) to show the correlation of the tumorigenesis and development of PTC. The PPI network had 23 nodes and 121 edges, with a local clustering coefficient of 0.67 and an average node degree of 9.68. The color reflected the degree of gene contribution, indicating whether there is a positive correlation with the occurrence of PTC. The nodes of MMP9, JUN, SPP1, HMOX1, and MMP1 were significantly enriched. The top 10 genes highly expressed in the PPI network were sorted by degree, and the specific data are shown in Table 1. The top 4 genes (MMP9, JUN, SPP1, and HMOX1) were selected based on the results for further experiments.

KEGG Pathway and GO Functional Enrichment Analyses.
The 23 commonly regulated genes were further analyzed to ascertain their biological processes. Biological processes with high enrichment scores included response to extracellular stimuli, oxidoreductase activity, and extracellular matrix organization (Figures 3(a)-3(c)). The results of KEGG (Kyoto Encyclopedia of Genes and Genomes) analysis suggested that these genes were enriched in tumor necrosis factor (TNF), nuclear factor kappa-B (NF-κB), and PI3K-AKT signaling pathways (Figure 3(d)). These data demonstrated that quercetin may function in PTC by these pathways.

Molecular Docking Analysis.
The results of molecular docking indicated the binding among quercetin and hub proteins, which may explain the mechanism of quercetin treatment in PTC in some ways. The molecular docking results showed that quercetin binds to central proteins, which may explain the mechanism of quercetin action on PTC to some extent. Quercetin inhibited the occurrence and progression of PTC by modulating MMP9, JUN, SPP1, and HMOX1 targets, which was in accordance with the results of NP screening. Molecular docking analysis validated the reliability of NP analysis. The results of quercetin docking with PTC target protein receptor sites are shown in Figure 4, and the top 10 molecular binding energies of quercetin docking with target proteins are shown in Table 2.

Discussion
PTC is the most common pathological type of thyroid malignancy, accounting for approximately 80-90%. The role of microenvironment in thyroid carcinogenesis has been 2 Disease Markers clarified [15]. The components of microenvironment (extracellular matrix components (ECM) and stromal cells) surround and support cancer cells, interacting with them through cell-cell and cell-extracellular matrix interactions [16]. These are consistent with the results we obtained; through GO analysis, we concluded that the role of quercetin in PTC is enriched in processes such as response to extracellular stimulus, oxidoreductase activity, and extracellular matrix organization. The poor outcomes of cancers are partially caused by the interaction of the tumor with the surrounding immune microenvironment and the collective effect of these interactions during tumorigenesis and progression, promoting through immune tolerance and immune escape, leading to the body's incapacity to completely clear the tumor [17].
Extensive epidemiological evidence suggests that fruits and vegetables may prevent some human cancers and be related to a reduced risk of cancer-related mortality [18][19][20][21][22]. Anticancer and anti-inflammatory activities are a range of the main quercetin mechanisms of action [23][24][25]. Previous experiments have shown that quercetin prevents the growth of multifarious tumors, including ovarian, breast, colon, and gastric tumors, by suppressing the cell cycle and inducing apoptosis; moreover, the mechanism cell signaling pathways might involve PI3K and MAPK pathways [26]. Quercetin has been reported to increase expression of differentiation marker NIS in dedifferentiated/anaplastic thyroid cancer cells [27]. A study found that combined utilization of quercetin and sorafenib induced inhibition of cell growth and contributed to reduce PTC cell motility, with a   Disease Markers significant increase of mRNA levels of E-cadherin expression and decrease of N-cadherin [28]. Another research found that quercetin suppressed the expression of Hsp90 and inhibited the growth of PTC cells, but the molecular mechanism of antiproliferation and apoptosis of Hsp90 was unclear [29].
Our results suggested that 23 target genes were regulated by quercetin in PTC cells. MMP9, JUN, SPP1, and HMOX1, with the top 4 degrees through PPI network analysis, were considered as the key genes. Matrix metalloproteinase 9 (MMP9), a key regulator of the extracellular matrix, has been regarded as an attractive therapeutic target, which participates in the degradation of many extracellular matrix proteins and also participates in inflammatory and oncological disorders [30]. MMP9 is critical for the remodeling of extracellular matrix components of the basement membrane such as type IV collagen and laminin [24]. Moreover, MMP9 regulates other cellular processes, and its expression and

Disease Markers
JUN, a key component of dimeric transcription factor activator protein-1 (AP-1), plays a role in the transcriptional response to extracellular signaling [37]. A previous study showed that AP-1 expression in papillary thyroid carcinoma cancer tissue is positively correlated with a high incidence of disease progression, a high tumor recurrence rate, and unfavorable disease-free survival [38]. In our experimental group, JUN expression was downregulated in PTC cells after  Disease Markers quercetin treatment. JUN might be used as a therapeutic molecular target to benefit a cure for PTC patients [39]. Secretory phosphoprotein 1 (SPP1), also known as bone bridge protein (OPN), is an integrin-binding glycophosphoprotein and has been shown to be overexpressed in many cancers such as hepatocellular carcinoma, lung cancer, prostate cancer, and breast cancer [40][41][42]. A study reported that SPP1 mRNA overexpressed in PTC samples [43] and the intensity of SPP1 staining was corresponding to the tumor size and lateral cervical lymph node metastasis (LNM) in human PTC [44]. Moreover, the level of SPP1 produced by macrophages has been shown to be related to the development of psammoma bodies (PBs) in PTC [45], which may be linked with the development of thyroid cancer. Our experiments showed that quercetin also regulates the expression of SPP1 in PTC.
In cancer cells, heme oxygenase-1 (HMOX1/HO-1) may improve cancer aggressiveness and resistance to therapy, resulting in poor prognosis [46]. Previous studies showed that HMOX1 was associated with advanced tumor stage in thyroid cancer [47].
According to the GO enrichment analysis, we found that biological process was focused on regulation of cell motility, cell adhesion mediated by integrin and response to extracellular stimulus including radiation, reactive oxygen species. Cellular component is enriched in extracellular matrix and membrane structure. As previously mentioned, MMP9 is a key regulator of the extracellular matrix and JUN plays a role in the transcriptional response to extracellular signaling; SPP1 is an integrin-binding glycophosphoprotein and HMOX1 overexpressed in cells such as macrophages supporting the immediate surroundings of the tumor [48], and the antioxidant role of HMOX1 in malignancies is well known [49].
Tumor-related pathways, such as PI3K-Akt, TNF, and NF-κB signaling pathways, may be involved in quercetin treating PTC through KEGG enrichment analysis. Gene alterations caused by partners in the PI3K-AKT signaling pathway are closely associated with thyroid cancer progression [50]. NF-κB is a key regulator of thyroid cancer angiogenesis [51]. For many years, it has been thought that the NF-κB pathway in thyroid cancer cells plays a protumor role [52].
Our study suggested that quercetin may suppress PTC through the above mechanisms, which are closely related to key target proteins. The tumor signaling pathways and biological processes were described in the study. Further experimental and clinical validation of our findings is needed.

Screening Differentially Expressed Genes (DEGs) in PTC.
From the Gene Expression Omnibus online database (https:// www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE33630), we obtained the dataset (GSE33630) of gene expression in normal thyroid tissue and PTC. The results were assessed by utilizing the limma and pheatmap packages in R software.

Protein-Protein Interaction (PPI) Network Construction.
The PPI network was constructed by utilizing the STRING 11.0 database (https://www.string-db.org/), and the mechanisms of these target protein interactions were visualized by Cytoscape 3.8.2 software. We entered the candidate target proteins into STRING to construct a PPI network; the minimum interaction score > 0:4. In the network, edge represented the PPI, while node represented target protein.

KEGG and GO Enrichment
Analyses. KEGG pathway analysis and GO enrichment analysis were performed by utilizing R software. The topological analysis targets were imported; KEGG enrichment analysis and MF (molecular function), CC (cellular component), and BP (biological process) data were collected; weighting analyses were performed, and bubble diagrams were generated. The pathways concerned with PTC were determined by the clinical and pathological data; adjusted p values < 0.05.

Molecular
Docking. The MOL2 format of quercetin's chemical structure (Figure 2(a)) was acquired from the Analysis Platform (TCMSP, Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform) database and then converting the file to PDB format using the PubChem database (https://www.pubchem.org/). The RCSB PDB database (http://www.rcsb.org/) was searched to obtain the protein crystal structure, and then, after dehydration and hydrogenation in AutoDock software, the structure was saved in PDB format. PyMOL software was used to analyze the quercetin-target proteins.        . The values of all counts were expressed as the mean standard deviation (SD) of 3 independent replicates. Comparative differences were tested using the Student t-test, with p < 0:05 and p < 0:01 indicating statistically significant differences.

Conclusions
In our study, quercetin systematically alters cell proliferation, invasion, and apoptosis of PTC cells by regulating the expression of several proteins such as MMP9, JUN, SPP1, and HMOX1. Through GO and KEGG enrichment analyses, we hypothesized that these protein molecules are mainly enriched in PI3K-Akt, NF-κB, and TNF pathways. The study confirmed the anticancer effects and potential targets of quercetin in PTC.

Data Availability
The data that support the findings of this study are openly available in https://www.ncbi.nlm.nih.gov/geo/query/acc