Anticolon Cancer Targets and Molecular Mechanisms of Tao-He-Cheng-Qi Formula

Background Tao-He-Cheng-Qi Formula (THCQF) is a traditional Chinese medicine that has been proven to have antitumor effects. The aim of this study was to elucidate the molecular targets and mechanisms of THCQF against colon cancer and construct a prognostic model based on network pharmacology, bioinformatics analysis, and in vitro experiments. Methods Potential THCQF compounds and targets were retrieved from the Traditional Chinese Medicine Systems Pharmacology and Bioinformatics Analysis Tool for Molecular Mechanism of Traditional Chinese Medicine databases. Differentially expressed genes for colon cancer were screened in The Cancer Genome Atlas and Gene Expression Omnibus databases. The anticolon cancer mechanisms of THCQF were explored using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Molecular docking simulations and molecular dynamics analysis were used to evaluate the binding between target proteins and active compounds. Finally, the identified compounds were used to treat colon cancer cells from the HCT116 cell line, and expression of mRNA and protein after relevant posttreatment were tested using real-time polymerase chain reaction and western blotting. Results A total of 27 anticolon cancer targets of THCQF were selected, among which four genes (CCNB1, CCNA2, IL1A, and MMP3) were shown to effectively predict patient outcomes in a prognostic colon cancer model. GO and KEGG enrichment analyses indicated that the activity against colon cancer of THCQF was associated with the interleukin (IL)-4 and IL-3 signaling pathways. Two compounds in THCQF, aloe emodin (AE) and quercetin (QR), were shown to efficiently bind to cyclin B1, the protein encoded by CCNB1. Finally, incubation of HCT116 cells with AE and QR significantly decreased CCNB1 mRNA expression and cyclin B1 levels. Conclusions Taken together, the results indicate that AE and QR are the pivotal active compounds of THCQF, and CCNB1 is the main molecular target through which THCQF exerts its anticolon cancer effects. The study findings provide insight for studies investigating the anticancer effects of other traditional Chinese medicines.


Background
Colon cancer is a ubiquitous gastrointestinal malignancy that accounts for 6% of new cancer cases and 5.8% of cancer deaths annually in the world [1]. e incidence of colon cancer has increased yearly owing to improved economic standards, which have altered the diet structure and habits of many populations [2]. Patients with early-stage colon cancer may not exhibit obvious symptoms, but they gradually develop altered bowel habits, fecal characteristics, and abdominal pain [3]. By the time patients with colon cancer seek treatment, 25% of cases experience tumor metastasis, which results in poor treatment outcomes [4]. Despite advances in colon cancer treatment, including surgery, chemotherapy, and radiotherapy [5], the survival rate of 5-year colon cancer patients with stage IV remains less than 12% [6,7]. erefore, more effective therapeutic approaches for colon cancer are urgently needed.
e formulation contains 12 g TR, 12 g DH, 6 g GZ, 12 g GC, and 6 g MX. THCQF has been used in East Asia to treat interior heat patterns and resolve blood stasis for more than a thousand years [9]. It has also been applied to treat acute liver failure [10], chronic kidney disease [11], and intracerebral hemorrhage [12]. Most importantly, THCQF has an inhibitory effect against cancers [13]. For example, THCQF inhibited the expression of BCL-2 and PCNA to suppress cancer development [14]. Furthermore, THCQF enhanced human immunity against cancer by promoting TNF-α expression [15]. Previous studies have shown that THCQF treatment can help alleviate the discomfort of patients with colon cancer [16] and prevent disease recurrence [17]. Nevertheless, the molecular targets and mechanisms of THCQF against colon cancer remain unclear and warrant further elucidation.
Tumor cells are characterized by various genetic changes and chromosomal instabilities that can be identified by bioinformatics analysis, which may enable a personalized approach to cancer treatment and help identify new drugs that target specific proteins [18]. e network pharmacology approach, which reveals druggene-disease comodule associations, has been widely used to predict target profiles and pharmacological effects of herbal compounds [19]. However, the effectiveness of predicted targets needs to be verified. erefore, in addition to exploring the anticolon cancer molecular targets and mechanisms of THCQF using network pharmacology and bioinformatics analysis, our study aimed to construct a prognostic model of survival status and validate the effects of the identified active compounds on the predicted targets in colon cancer cells.

Screening of THCQF Candidate Compounds and Targets.
e workflow of this study was shown in Figure 1. e major candidate compounds of THCQF were obtained from the TCM systems pharmacology (TCMSP) database (https://lsp.nwu.edu.cn/tcmsp.php) using the following criteria: oral bioavailability ≥30% and drug likeness ≥0.18. In the TCMSP database, relationships between Chinese herbal medicines, targets, and diseases are captured. Chemicals, targets, and drug-target networks are included, along with pharmacokinetic properties such as oral bioavailability, drug likeness, intestinal epithelial permeability, blood-brain barrier in this database [20]. Because the TCMSP database lacks information on minerals, the major compounds of MX were screened in the Bioinformatics Analysis Tool for Molecular Mechanism of TCM (BATMAN-TCM) database (https://bionet. ncpsb.org/batman-tcm/) using the following criteria: score cutoff ≥20 and adjusted P value < 0.05. e BATMAN-TCM database enables predictions of potential targets for each query ingredient, and performs Gene Ontology (GO) term and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses for target compounds.
ese functions contribute to the understanding of multicomponent, multitarget, and multipathway therapeutic mechanisms of TCMs and provide clues for further experimental validation [21]. Perl software was performed to create a candidate compound and target network.

Identification of erapeutic Targets of THCQF against
Colon Cancer. Transcriptome data for colon cancer and associated metadata were downloaded from e Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer. gov/). Gene expression profiles from GSE68468, a colon cancer dataset, and GPL96, the Affymetrix Human Genome U133A Array dataset, were downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi. nlm.nih.gov/geo/). e "limma" package in R software was applied to detect differentially expressed genes (DEGs) between normal and tumor groups using the following criteria: log|fold change| > 1, adjusted P value <0.05, respectively.
e THCQF targets were integrated with the DEGs to identify therapeutic targets of THCQF against colon cancer.

Construction of a Prognostic Model of Colon Cancer.
Univariate Cox regression analysis was employed to screen for anticolon cancer targets of THCQF that were significantly associated with patient survival. Least absolute shrinkage and selection operator (LASSO) regression analysis was used to remove redundant factors and build a prognostic model. e LASSO-Cox model has been widely applied for cancer survival prediction and detecting associated genes because it is capable to identify these genes that with high internal consistency [22,23]. Kaplan-Meier (KM) survival analysis was used to divide the risk scores of the model into high-and low-risk groups, with a P value <0.05 in the log-rank test considered to be significant. In addition, receiver operating characteristic curves were adopted to evaluate the predictive capability of the model for 1-, 3-, and 5-year survival. An area under the curve (AUC) > 0.5 was considered to indicate good predictive capability.

Construction of a Protein-Protein Interaction (PPI) Network and Identification of Key Targets.
To identify the hub target genes of THCQF against colon cancer, a PPI network was constructed using Metascape (https://metascape.org/gp/ index.html). Targets in the network were clustered based on their interaction relationships, and those with similar functions were grouped into modules. e genes selected for the prognostic model were overlapped with the targets of the PPI network to identify the hub targets that THCQF anticolon cancer.  Evidence-Based Complementary and Alternative Medicine survival-associated genes were considered core candidates for further research.

GO and KEGG Functional Enrichment Analyses of erapeutic Targets of THCQF against Colon
Cancer. GO and KEGG functional enrichment analyses of the therapeutic targets of THCQF against colon cancer were performed using Metascape according to the criteria: P value <0.01 and minimum enrichment >1.5. Additionally, an interaction network was constructed to explore possible connections between the identified biological functions and pathways.

Molecular Docking
Simulation. Molecular docking simulations were performed to verify the binding abilities of the active small molecules to the identified proteins. e three-dimensional protein structures were downloaded from the Protein Data Bank (https://www.rcsb.org/). Subsequently, hydrogens, charges, and protons were added to the proteins using the optimized potentials for liquid simulation (OPLS) 2005 force field and energy optimization was carried out using the "Protein and Ligand Preparation" module in Schrödinger software. e "Glide" docking method in Schrödinger software was used to predict the binding modes of the small molecules and proteins.

Molecular Dynamics Analysis.
e stability of the predicted small molecule-protein complexes was assessed using Amber molecular dynamics v14 software. e small molecule-protein complexes predicted via molecular docking simulation were used as the initial structures. e general Amber force field (GAFF) was applied for ligands, and Amber ff14SB force field was used for proteins. e default protonation state was used for amino acid residues of the proteins, and hydrogenation was carried out using the "tleap" module. e optimized structures and electrostatic potential of the small molecule-protein complexes were calculated using the "Gaussian09" quantum mechanical software package with the density functional theory level B3LYP/6-31G. e partial charges of the small molecules were obtained using the restrained electrostatic potential method in the "Antechamber" module. A constant volume was used to heat the system from 0 to 300 K within 60 picoseconds (ps), and the solvent density was balanced in a constant temperature-pressure system (T � 300 K, P � 1 atm). Sampling was set to 100 nanoseconds under constant pressure conditions with one frame per ps for subsequent analysis of conformation. e stability of the protein was evaluated using the root-mean-square deviation (RMSD), while the flexible region of the protein was assessed by the root-mean-square fluctuation (RMSF).

Cell Viability Assay.
To verify the anticolon cancer activity of the identified compounds, HCT116 colon cancer cells, which were purchased from the Chinese Academy of Sciences Cell Bank and routinely cultured in DMEM/F12 medium containing 10% fetal bovine serum in a saturated humidity incubator with 5% CO 2 at 37°C, were incubated with different concentrations of candidate compounds for 72 h, and the Cell Counting Kit-8 assay was performed to assess cell viability. e half-maximal inhibitory concentration (IC 50 ) of each compound against HCT116 cells was calculated. Subsequently, HCT116 cells were incubated with the IC 50 of each compound for 12 h, and the relevant mRNA and protein expression levels were evaluated using real-time polymerase chain reaction (RT-PCR) and western blotting, respectively. Aloe emodin (AE; Lot 210716, purity >98%) and quercetin (QR; Lot 210621, purity >98%) were obtained from Chengdu PureChem-Standard Co., Ltd. (Chengdu, China).

RT-PCR.
Reverse transcription was carried out using an RT-PCR kit (Novizan Biotechnology Co., Ltd., Nanjing, China) according to the manufacturer's instructions with 1 μg total RNA. e reaction was prepared as shown in Table 1. After reverse transcription, the reaction mixture was diluted 1 : 10 with ddH 2 O and stored at −20°C. Subsequently, RT-PCR was carried out in a 96-well plate using a real-time fluorescence quantitative PCR instrument (CFX96 Bio-Rad) according to the manufacturer's instructions. e specific primers are listed in Table 2, and GAPDH was used as the internal reference for normalization of gene expression. e 2 −ΔΔCt method was used for relative quantification of gene expression. Each compound was tested in triplicate.

Western
Blotting. Samples with equal amounts of protein, along with molecular weight standards, were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis at 100 V for 1-2 h. e separated proteins were transferred to a polyvinylidene fluoride membrane at a constant current of 200 mA for 1 h. e membrane was incubated with 5% blocking buffer at room temperature for 1 h, followed by incubation with a primary antibody with a concentration 1 : 1000 at an appropriate dilution at room temperature for an additional 2 h. After that, the membrane was washed 3 times successively with TBST (Tris-buffered saline with Tween 20) for 5 minutes. Subsequently, the membrane was incubated for 1 h at room temperature in TBST containing 5% blocking reagent and a labeled secondary antibody with a concentration 1 : 5000 and then washed three times with TBST for 5 min each. Finally, the membrane was treated with an enhanced chemiluminescence reagent to visualize the protein bands. GAPDH was used as the internal reference for protein expression, and the relative protein expression in each band was quantified using Quantity One software (Bio-Rad Laboratories, Hercules, CA, USA).

Identification of THCQF Candidate Compounds and
erapeutic Targets against Colon Cancer. e candidate THCQF compounds identified in the TCMSP and BAT-MAN-TCM databases included 16 compounds from DH, 92 from GC, 7 from GZ, 1 from MX, and 23 from TR. A total of Type Type  Up to 20 Type Type  Evidence-Based Complementary and Alternative Medicine 184 targets of THCQF were obtained by mapping the candidate ingredients and potential genes. ere were 3,481 and 1,602 DEGs were screened out in the TCGA database (Figures 2(a) and 2(b)) and GSE68468 dataset (Figures 2(c) and 2(d)), respectively. By integrating the THCQF targets with the DEGs, 27 therapeutic targets of THCQF against colon cancer were finally identified (Figure 2(e)).

Construction of a Prognostic Model of Colon Cancer.
Univariate Cox regression analysis identified four genes (CCNB1, CCNA2, IL1A, and MMP3) that significantly affected patient survival based on the 27 anticolon cancer targets of THCQF. Redundant factors were removed by LASSO regression analysis and CCNB1, CCNA2, IL1A, and MMP3 were used to build the prognostic model (Figures 3(a) and 3(b)). KM survival analysis revealed a  Evidence-Based Complementary and Alternative Medicine significant difference between high-and low-risk groups (P � 0.0196 in the log-rank test) (Figure 3(d)). e median survival time was 5.4 years in the high-risk group and 7.1 years in the low-risk group. Furthermore, the AUC values were 0.566, 0.575, and 0.581 for 1-, 3-, and 5year survival, respectively, suggesting that the prognostic model effectively predicted the survival of patients with colon cancer (Figure 3(e)).

Construction of PPI and Molecule-Target Networks and
Identification of Key Targets. e PPI network was built up using 27 therapeutic targets of THCQF against colon cancer (Figure 4(a)). Cluster analysis finally identified two key functional modules with nine therapeutic targets, including CCNB1, CCNA2, MYC, E2F1, CDKN1A, CCND1, CHEK1, TOP2A, and PRKCB (Figure 4(b)). Overlapping the four genes used in the prognostic model with the nine targets of    Evidence-Based Complementary and Alternative Medicine the PPI network yielded two key anticolon cancer targets, namely, CCNB1 and CCNA2. A molecule-target network was developed to screen out the pivotal candidate ingredients of THCQF with anticolon cancer activity using the 139 compounds and 27 therapeutic targets of THCQF ( Figure 5). Within the network, two compounds from DH, four from TR, one from MX, one from GZ, and twenty-five from GC were identified to potentially possess anticolon cancer activity (Table 3). Notably, CCNB1 was targeted by one compound each from DH and GC, while CCNA2 was targeted by twenty compounds from GC.

GO and KEGG Functional Enrichment Analyses.
GO enrichment analysis was employed to examine the functions of biological systems of the 27 therapeutic targets. e results indicated that these genes were involved in oxygen level response and lipid metabolism. In the meantime, KEGG pathway enrichment indicated that the major mechanisms of THCQF inhibited colon cancer were associated with the interleukin (IL)-4 and IL-3 signaling pathways (Figure 6).

Molecular Docking and Molecular Dynamics of Target Proteins and Candidate Compounds.
e molecular docking results indicated that the binding affinity of AE and QR to cyclin B1 was good, the protein encoded by CCNB1, compared to other compound-protein complexes (Table 4). Notably, AE had the top-ranked docking score of −6.124 with cyclin B1. erefore, AE and QR were selected for further investigation (Figure 7). e molecular dynamics results indicated that although the RMSD trajectories of AE and QR bound to cyclin B1 displayed small fluctuations, they remained stable overall.
e RMSD values suggested that AE and cyclin B1 formed a more stable complex than QR and cyclin B1, indicating a higher binding affinity between AE and cyclin B1 (Figures 8(a) and 8(b)). Meanwhile, the RMSF values of the AE-bound amino acid residues of cyclin B1 did not significantly fluctuate during the simulation process, indicating that the protein structure was more stable than that formed with QR (Figure 8(c)). e free energy of binding was calculated based on the conformation obtained by the molecular dynamics simulation based on the molecular mechanics/Poisson-Boltzmann surface area method. e total binding free energy (ΔGtot) of AE and cyclin B1

Validation of Anticolon Cancer Activity of AE and QR.
Colon cancer cell growth was not inhibited by AE and QR at concentrations below 25 μg/mL. However, cell growth was significantly inhibited in a concentration-dependent manner at compound concentrations between 25 and 800 μg/mL (Figures 9(a) and 9(c)). Notably, AE had a lower IC 50 than QR (41.29 ± 2.55 vs. 47.48 ± 3.59 μg/mL, resp.), which concurred with the molecular docking simulation and molecular dynamics analysis results (Figures 9(b) and 9(d)).
To investigate the molecular mechanisms of AE and QR against colon cancer, HCT116 cells were treated with AE and QR at their IC 50 values, and CCNB1 expression was examined. Both CCNB1 mRNA expression and cyclin B1 levels were significantly decreased after AE and QR treatment, further confirming that these two THCQF compounds targeted CCNB1 to exert anticolon cancer effects (Figures 9(e)-9(h)).

Discussion
Colon cancer is a common gastrointestinal malignancy with high incidence and mortality, posing a heavy socioeconomic burden [24]. Presently, colon cancer is treated primarily through surgical resection, in combination with chemotherapy and radiotherapy. However, adverse reactions caused by these treatments seriously affect patient quality of life [25]. THCQF is a frequently used TCM against lower energizer heat patterns and blood stasis [26] and is often used as an adjuvant therapy in clinical practice owing to its significant antitumor effects [27]. Emodin, derived from DH, has been confirmed to suppress the growth and invasive ability of SW480 colon cancer cells via the mitogen-activated protein kinase signaling pathway [28]. Aqueous GZ extract exerts an anticancer effect by targeting tubulin proteins, thus impeding mitosis and promoting cell apoptosis [29]. Research has demonstrated that the active compounds of GC may inhibit the proliferation and migration of colorectal cancer cells by blocking the S phase of the cell cycle [30]. Moreover, the application of THCQF has significantly  reduced postoperative inflammation, the occurrence of adverse reactions, and restored gastrointestinal function after colon cancer surgery [16]. However, the molecular targets, mechanisms of action, and active ingredients in THCQF have not yet been elucidated.
In the current study, 27 therapeutic targets of THCQF against colon cancer were obtained across multiple databases, and GO and KEGG functional enrichment analyses suggested that the effects of anticolon cancer was closely associated with biological processes such as oxygen level response and lipid metabolism, as well as the IL-4 and IL-3 signaling pathways. IL-4, a multifunctional cytokine produced by T helper 2 cells, is involved in regulation of the cell cycle, cell differentiation, and migration, and has an important effect on immune function [31,32]. In the early stage of tumor formation, an abnormal immune response caused by IL-4 promotes cancer cell mutation and DNA damage, thus promoting the occurrence and development of tumors [33]. Additionally, the secretion of IL-4 influences the tumor microenvironment by activating M2 macrophages, which leads to tumor immune escape, invasion, and metastasis [34]. IL-3, one of the most common inflammatory growth factors, can stimulate the proliferation and differentiation of myeloid elements [35], activate basophils to participate in the immune response [36], and inhibit angiogenesis by increasing TGF activity early in tumor formation [37]. Furthermore, induction of IL-3 expression in the bone marrow microenvironment of patients with myeloma has been shown to increase bone destruction and promote tumor cell growth [38]. e above results support that the anticancer effects of THCQF may be mediated via the IL-3 and IL-4 signaling pathways.
Four prognostic genes were selected among the 27 targets to construct an effective prognostic model of colon cancer. Two core targets (CCNB1 and CCNA2) were obtained by overlapping the PPI network targets with the four genes used in the prognostic model. CCNB1 is an important regulator of the cell growth cycle and can affect a wide range of biological processes, including inhibition of cell proliferation, induction of cell senescence, and apoptosis [39]. CCNB1 is widely expressed in a variety of tumors, such as ovarian cancer [40], liver cancer [41], and gastric cancer [42]. A previous study reported that CCNB1 inhibited the activation of the p53 pathway via regulation of FOXM1 to promote the occurrence and development of tumors [43]. Further research demonstrated that CCNB1 was continuously downregulated after knocking out SOX2OT, which resulted in inhibited migration and invasion of colorectal cancer cells due to reduced expression of the mesenchymal protein N-cadherin and increased expression of the epithelial protein E-cadherin [44]. ese findings suggest that CCNB1 may play an oncogenic role in colon cancer.
Furthermore, the compound-target network of CCNB1 revealed that multiple compounds targeted cyclin B1, including AE from DH and QR from GC. As an anthraquinone and TCM monomer, AE has a wide range of pharmacological effects, including antitumor activity [45], cardiovascular protection [46], anti-inflammatory activity  [47], and immune function regulation [45]. AE antitumor activity has been reported against liver cancer [48], breast cancer [49], and gastric cancer [50]. In SW260 and HT29 colorectal cancer cells, AE treatment induced the production of reactive oxygen species and endoplasmic reticulum stress, thus inhibiting cell viability and inducing cell apoptosis [51]. QR, a natural flavonoid widely distributed in nature, possesses a variety of biological activities and exerts antioxidant, antibiosis, and antitumor effects [52]. QR treatment reportedly induced the overexpression of TP53I3 and RRM2 via p53 signaling, resulting in increased apoptosis and inhibiting the proliferation and metastasis of cancer cells [2]. Research has also demonstrated that QR promoted the expression of caspase-3 and increased the BAX/BCL-2 ratio to induce apoptosis in human HT-29 colon cancer cells [53]. Taken together, these studies support the notion that AE and QR may play antitumor roles by targeting CCNB1.
Molecular docking simulations and molecular dynamics analysis provided further evidence supporting the predicted active compounds and target proteins, indicating that AE and QR had good binding affinities for cyclin B1. e in vitro results further confirmed that AE and QR inhibited the proliferation of HCT116 colon cancer cells and significantly reduced CCNB1 expression. As CCNB1 is known to be a tumor promoter that influences the immune activity of the hepatocellular carcinoma tumor microenvironment [54], these results suggest that AE and QR inhibit the proliferation of colon cancer cells by downregulating CCNB1.

Conclusions
In conclusion, this study explored the molecular targets and mechanisms of THCQF against colon cancer and constructed a relevant prognostic model using network pharmacology and bioinformatics analysis. e predicted molecular targets and active compounds of THCQF were further validated experimentally. Our findings suggest that the anticolon cancer activity of THCQF may be achieved via AE and QR targeting CCNB1. In addition to elucidating the novel molecular mechanism by which THCQF exerts its anticolon cancer effects, this study provides a theoretical basis for studying the anticancer mechanisms of other TCMs.