Multiomics Study of a Novel Naturally Derived Small Molecule, NSC772864, as a Potential Inhibitor of Proto-Oncogenes Regulating Cell Cycle Progression in Colorectal Cancer

Colorectal cancer (CRC) is one of the most prevalent malignant tumors, and it contributes to high numbers of deaths globally. Although advances in understanding CRC molecular mechanisms have shed significant light on its pathogenicity, current treatment options, including combined chemotherapy and molecular-targeted agents, are still limited due to resistance, with almost 25% of patients developing distant metastasis. Therefore, identifying novel biomarkers for early diagnosis is crucial, as they will also influence strategies for new targeted therapies. The proto-oncogene, c-Met, a tyrosine kinase that promotes cell proliferation, motility, and invasion; c-MYC, a transcription factor associated with the modulation of the cell cycle, proliferation, apoptosis; and cyclin D1 (CCND1), an essential regulatory protein in the cell cycle, all play crucial roles in cancer progression. In the present study, we explored computational simulations through bioinformatics analysis and identified the overexpression of c-Met/GSK3β/MYC/CCND1 oncogenic signatures that were associated with cancer progression, drug resistance, metastasis, and poor clinical outcomes in CRC. We further demonstrated the anticancer activities of our newly synthesized quinoline-derived compound, NSC772864, against panels of the National Cancer Institute’s human CRC cell lines. The compound exhibited cytotoxic activities against various CRC cell lines. Using target prediction tools, we found that c-Met/GSK3β/MYC/CCND1 were target genes for the NSC772864 compound. Subsequently, we performed in silico molecular docking to investigate protein–ligand interactions and discovered that NSC772864 exhibited higher binding affinities with these oncogenes compared to FDA-approved drugs. These findings strongly suggest that NSC772864 is a novel and potential antiCRC agent.


Introduction
Colorectal cancer (CRC) is the third most common malignancy worldwide and ranks as the fourth leading cause of cancer deaths [1,2]. Although remarkable progress has been achieved through current multimodality therapies, including chemotherapy and moleculartargeted agents, such as cetuximab [3][4][5][6][7], most patients eventually still become resistant to therapy [8]. Approximately 25% of CRC patients develop advanced metastatic cancer, which leads to poor clinical outcomes [9]. The 5-year overall survival rate in the initial stage is approximately 90%; however, survival of advanced-stage CRC patients is merely 10% [10]. Dysregulation of various genetic and molecular processes was reported in CRC, thus necessitating the urgent need to identify novel biomarkers, which can be used for early diagnosis and drug development [11,12]. The mesenchymal-epithelial transition (c-Met) is a tyrosine kinase receptor in the membrane encoded by the c-Met proto-oncogene [13]. The dysregulation of c-Met signaling was reported in CRC patients, and it promotes tumor angiogenesis, growth, and metastasis [14][15][16][17][18][19]. Studies showed that c-Met overexpression at the messenger (m) RNA and protein levels is linked to CRC progression, stemness, distant metastasis, and poor clinical outcomes [20][21][22].
Moreover, recent reports demonstrated that when activated in CRC patients, c-Met promotes resistance to cetuximab [23]. Therefore, elucidating the molecular mechanisms of c-Met overexpression in CRC pathogenesis would be of great significance for determining an effective approach for CRC therapy. Moreover, c-Met-dependent and -independent signaling contributes to Met receptor-mediated cell cycle progression [20]. The c-Met pathway has been shown to modulate various downstream signaling pathways in tumors, including PI3K/AKT and Wnt/β-catenin, among others [24,25]. The association of c-MET with Wnt/β-catenin in particular has been demonstrated to promote cancer proliferation, progression, and metastasis [26]. Several other studies have also shown that c-Met expression enhances the Wnt/β-catenin signaling and impedes GSK3β from phosphorylating β-catenin in CRC; this subsequently promotes β-catenin translocation into the nucleus leading to cancer initiation [27,28]. It has been shown that the inhibition of c-Met also suppresses Wnt pathway signaling transduction activities in cancer cells [29,30]. Glycogen synthase kinase 3 (GSK3), a serine/threonine kinase, and its expression in CRC associates with the regulation of cell cycle, cancer proliferation, and apoptosis [29], thus making it a potential biomarker for therapeutic development in CRC [31][32][33]. CRC is mainly known for its heterogeneity, including alterations in the expression of c-MYC, which was shown to be associated with CRC progression through the c-Met signaling pathway [34]. Accumulating studies have shown that MYC is a potential target of β-catenin signaling in CRC cells [35]. The c-MYC proto-oncogene is a transcription factor that modulates the cell cycle, proliferation, and apoptosis. Additionally [36,37], when upregulated in CRC, it is associated with cancer progression, recurrence, and metastasis [38][39][40][41][42][43][44]. The overexpression of c-MYC was shown to enhance the expression of cyclin D1 (CCND1) through activating beta-catenin [45], thus promoting CRC proliferation [46]. CCND1 is a crucial regulator of cell progression through the G 1 to S phases of the cell cycle via binding to cyclin-dependent kinase 4 (CDK4) and CDK6 [47].
Studies showed that CCND1 overexpression is linked to poor prognoses in CRC patients [48,49]. These findings suggest crosstalk among c-Met/GSK3β/MYC/CCND1 as a potential prognostic biomarker of CRC, which might be used to develop novel and effective therapeutics. In this study, we utilized a bioinformatics analysis to identify potential oncogenes, which are overexpressed in CRC Interestingly, our results reveal that c-Met/GSK3β/MYC/CCND1 were among the top 25 differentially expressed genes (DEGs) in CRC, extracted from the GEO microarray datasets, GSE4107 and GSE41328. MicroRNAs (miRNAs) are noncoding short RNAs (21-25 nucleotides) that negatively regulate gene expression at the translational level by binding to 3 UTR (untranslated region) of their target mRNAs leading to either their degradation or translational inhibition [50,51]. The dysregulation of expression levels in various miRNAs, including miR-26a, has been shown to associate with proliferation and invasion in different cancer types, including CRC [52]. miR-26a is significantly overexpressed in CRC. However, its function and clinical significance in CRC remain elusive [53]. Through computational simulation, we identified miR-26a bound to the 3 UTR of GSK3β/MYC/CCND1, thus suggesting a close association of Met/GSK3β/MYC/CCND1 and miR-26a promoting cancer proliferation, invasion, migration, drug resistance, and stemness metastasis in CRC patients.
The search for a novel and effective cancer therapy has increased over the years due to resistance to current treatment modalities. Quinoline is a heterocyclic aromatic nitrogen compound commonly used as a parental compound to synthesize various molecules [54]. Numerous studies demonstrated that several quinoline-derived molecules exhibit anticancer activities due to their ability to inhibit specific proteins [55]. One of the quinoline alkaloids, camptothecin, is an S-phase cell cycle-specific anticancer agent shown to induce topoisomerase I-dependent DNA breaks [56,57]. Moreover, sunitinib, a multitargeted receptor tyrosine kinase (RTK) inhibitor [58], which was approved as an anticancer agent in gastrointestinal stromal tumor patients, was shown to inhibit cancer cell growth expressing RTKs [59,60]. In the current study, we evaluated the anticancer effects of NSC772864, a novel small-molecule derivative of quinoline and sunitinib, which was recently developed in our lab [61][62][63][64][65][66]. Computational simulations are extensively used in drug discovery and development; therefore, we applied in silico predictions and screened our NSC772864 small molecule for its anticancer activities in a panel of National Cancer Institute (NCI) CRC cell lines to predict novel drug targets [67]. We further identified c-Met/GSK3β/MYC/CCND1 as druggable candidates for this compound and further performed molecular docking simulations of ligand-receptor interactions of NSC772864 in complex with c-Met/c-MYC/CCND1 study summary is shown in Figure 1.

Retrieval of the Top 25 DEGs in CRC
The top overexpressed DEGs in colon adenocarcinoma (COAD) samples were retrieved from the following databases: NCBI GEO two datasets, GEO microarray datasets, and GSE4107 and GSE41328 (https://www.ncbi.nlm.nih.gov/ 21 June 2022). We used ULCAN, a web-based tool to analyze The Cancer Genome Atlas (TCGA) level 3 RNA sequencing (RNA-Seq) and clinical data from tumor tissues compared to adjacent normal tissues [68]. The top overexpressed genes with significantly different transcripts per million (TPM) values (p < 0.001) were first selected and represented in a heatmap.

Retrieval of the Top 25 DEGs in CRC
The top overexpressed DEGs in colon adenocarcinoma (COAD) samples were retrieved from the following databases: NCBI GEO two datasets, GEO microarray datasets, and GSE4107 and GSE41328 (https://www.ncbi.nlm.nih.gov/ 21 June 2022). We used ULCAN, a web-based tool to analyze The Cancer Genome Atlas (TCGA) level 3 RNA sequencing (RNA-Seq) and clinical data from tumor tissues compared to adjacent normal tissues [68]. The top overexpressed genes with significantly different transcripts per million (TPM) values (p < 0.001) were first selected and represented in a heatmap.

DEG Validation
To validate expression levels of the identified DEGs, including c-c-Met/GSK3β/MYC/ CCND1 in CRC, we explored the tumor, normal, and metastatic plot tool (https://tnmplot. com/analysis/ 28 June 2022), which compares selected genes from RNA-Seq-based data [69]. A comparative analysis was performed using the Kruskal-Wallis test method, with p < 0.05 considered statistically significant. Expressions of c-Met/GSK3β/MYC/CCND1 based on individual cancer stages were analyzed by utilizing ULCAN software. Furthermore, we used GECO (https://proteinguru1.shinyapps.io/geco/ 28 June 2022), a gene expression correlation analytical tool, which differentiates two expression profiles into positive and negative correlations [70], with positive Pearson correlation coefficients and p < 0.05 as statistically significant.

Interacting Network Construction and Functional Enrichment Analysis
A protein interaction network (PIN) was analyzed using STRING, a protein-protein interaction (PPI) and functional enrichment analysis web tool (https://string-db.org/ 29 June 2022) [71]. We further downloaded functional enrichment clusters, from the STRING analysis, and these included Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and gene ontology (GO), which involves biological processes (BPs), molecular functions (MFs), and cellular components (CCs). Functional enrichments of the selected clustering network were further classified using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/ 29 June 2022) functional annotation tool [72]. Moreover, we explored the functional enrichment software database, FunRich (http://www.funrich.org/ 29 June 2022), to further analyze interactions within the networks, with p < 0.05 considered significant.

Analysis of c-Met/GSK3β/MYC/CCND1 Genetic Alterations and Mutation Analysis in CRC Tissues
To determine relationships between c-Met/GSK3β/MYC/CCND1 mutations and alterations in gene expressions, we utilized the muTarget platform (http://www.mutarget.com 06 August 2022) and web tool that link gene expression alterations and mutation statuses in cancers [73]. We further used the cBio Cancer Genomics Portal (cBioPortal; http://cbioportal.org 6 August 2022), an online interactive platform for translational research that provides analysis of cancer genomic datasets [74], to determine alteration frequencies of selected gene signatures in CRC tissues. Furthermore, based on the genetic alteration profile of the cohorts, each was categorized into one of two groups: an unaltered group and an altered group. The analytical results were considered significant at p < 0.05.

Drug Sensitivity and Gene Expression Profiling for c-Met/GSK3β/MYC/CCND1 Oncogenes in Colon Tumors
Correlations between expressions of c-Met/GSK3β/MYC/CCND1 oncogenes and drug sensitivity were determined using the GSCALite database, an online tool for Gene Set Cancer Analysis (GSCA) (http://bioinfo.life.hust.edu.cn/web/GSCALite/ 11 August 2022) [75]. The analysis was performed using the top 38 drugs in a pancancer database. Expression levels of individual genes in the gene set were analyzed using a Spearman correlation analysis in response to the drug sensitivity (50% inhibitory concentration (IC 50 )) of small molecules. Additionally, correlations considered statistically significant displayed a false discovery rate (F.D.R.) of <0.05. The Developmental Therapeutics Program (D.T.P.)-COMPARE online platform, from the NCI (https://dtp.cancer.gov/ 15 June 2022) is an online pattern recognition algorithm used to predict molecular targets based on ARRAYCGH_GRAY and detect a compound's similarities to NCI synthetic compounds and standard agents. Herein, we used 50% growth inhibition (GI 50 ), which is an endpoint of the IC 50 , and the National Safety Code (N.S.C.) number (772864) as delimiters [76]. Furthermore, we used the PharmMapper Server (http://59.78.96.61/pharmmapper 14 August 2022), an integrated pharmacophore mapping platform, to identify potential drug target candidates [77]. Identifying novel and potential drug candidates early in drug discovery and development is crucial, as it reduces time and costs; herein, we applied the drug-likeness concept based on specific criteria [78,79]. To evaluate the drug-likeness, pharmacokinetics (PKs), and medicinal chemistry of the NSC772864 compound, we utilized swissADME (http://www.swissadme.ch/ 15 June 2022), an online platform used to evaluate PKs, druglikeness, and the medicinal chemistry friendliness of compounds [80]. We analyzed the drug-likeness properties according to the Lipinski (Pfizer) rule of five, Ghose (Amgen), Veber (GSK), and Egan (Pharmacia) standards and further showed relationships between the PKs and physicochemical properties [81]. Moreover, we analyzed the gastrointestinal absorption (GIA) and blood-brain barrier (BBB) penetration properties using the brain or intestinal estimated permeation (BOILED-Egg) model [82]. The Abbot bioavailability score was determined based on the probability of the compound having at least 10% oral bioavailability in rats or measurable Caco-2 permeability [83].

In Vitro Anticancer Screening of NSC77286 against NC1 60 CRC Cells
NSC772864 was submitted to the NCI Developmental Therapeutics Program (DTP) for screening for potential antiproliferative and cytotoxic effects against a panel of NCI 60 CNS cell lines, in agreement with the protocol outlined by the NCI (https://dtp.cancer.gov/ 30 June 2022). The compound was tested at an initial dose of 10 µM and further tested for dose-dependent treatment. The results show that NSC772864 exhibited antiproliferative activities against several CNS cell lines [84].

High Expressions of c-Met/GSK3β/MYC/CCND1 Promote Colon Cancer Progression
The top overexpressed DEGs in COAD tissues compared to normal tissues were retrieved from two NCBI GEO datasets, GSE4107 and GSE41328 (Figure 2A,B). We used a Venn diagram to demonstrate the overlapping upregulated genes from the two datasets ( Figure 2C). Subsequently, we analyzed these DEGs using the ULCAN database to construct the heatmap with the overexpressed genes. The c-Met/GSK3β/MYC/CCND1 oncogenic signature was identified among the top-ranking DEGs. The gene expression level is represented as log2(TPM +1) ( Figure 2D). The TNM plots revealed a high expression of c-Met/GSK3β/MYC/CCND1 in tumor tissues compared to normal colon tissues ( Figure 2E

PPI Network (PIN) Construction and Enrichment Analysis
PPIs were analyzed using the STRING database, and the analytical results show an average local clustering coefficient of 0.920 and interaction enrichment with p = 0.0135. Interactions were analyzed based on coexpression, gene fusion, and co-occurrence ( Figure  3A). Functional enrichments with the clustering network were retrieved from the STRING analysis, and they included GO involving BPs, which were analyzed using FunRich software. Among the top 10 enriched BPs were E-cadherin signaling, the FOXM1

PPI Network (PIN) Construction and Enrichment Analysis
PPIs were analyzed using the STRING database, and the analytical results show an average local clustering coefficient of 0.920 and interaction enrichment with p = 0.0135. Interactions were analyzed based on coexpression, gene fusion, and co-occurrence ( Figure 3A). Functional enrichments with the clustering network were retrieved from the STRING analysis, and they included GO involving BPs, which were analyzed using FunRich software. Among the top 10 enriched BPs were E-cadherin signaling, the FOXM1 transcription factor network, G 1 phase, cyclin D, and mitosis G 1 -G 1 /S phases ( Figure 3B). For further analysis, we used the DAVID database to evaluate enriched KEGG pathways and CCs. Interestingly, among the top 15 enriched pathways were MYC, CCND1, lysine acetyltransferase 2A (KAT2A), sirtuin 1 (SIRT1), S-phase kinase-associated protein, c-Met, and hepatocyte growth factor, while the enriched CCs included CCND1, cyclin-dependent kinase 4 (CDK4), and KAT2A ( Figure 3C,D), with p < 0.05 considered significant.

Crosstalk between Overexpression of Met/GSK3β/MYC/CCND1 Oncogenes and Upregulated miR-26a are Associated with CRC Cancer Progression
Migration and metastasis play a significant role in determining the clinical outcomes of colorectal cancer (CRC). A major factor for metastasis is the acquired capacity of the cell

Crosstalk between Overexpression of Met/GSK3β/MYC/CCND1 Oncogenes and Upregulated miR-26a Are Associated with CRC Cancer Progression
Migration and metastasis play a significant role in determining the clinical outcomes of colorectal cancer (CRC). A major factor for metastasis is the acquired capacity of the cell to proliferate and invade normal tissues. Herein, we investigated the role of miR-26a in regulating CRC cell proliferation and migration, through the crosstalk with upregulated Met, GSK3β, MYC, CCND1, and oncogenic signatures [53]. We used the FunRich and Net-workAnalyst comprehensive gene and miRNA expression profiling platforms [89,90], and profiled all the miRNAs that targets c-Met, GSK3β, MYC, and CCND1 oncogenes specifically ( Figure 4). For further analysis, we identified OncomiR-26a to be enriched within the same clustering network and linking with Met/GSK3β/MYC/CCND1 oncogenes ( Figure 5A), and to explore further, we used the ENCORI tool and identified miR-26a to be downregulated in CRC samples as compared to normal samples ( Figure 5B). Moreover, we identified the miR-26a-1-3p site in the 3 untranslated region (UTR) of GSK3β, MYC, and CCND1 mRNA, thus predicting a link between miR-26a and Met/GSK3β/MYC/CCND1 in cancers ( Figure 5C,D). To explore further, we performed an enrichment analysis and identified all the enriched gene ontology including the biological process and the affected KEGG pathways ( Figure 5F,G).

Genomic Alterations in c-Met/GSK3β/MYC/CCND1 Signatures Are Associated with Poor Prognoses of CRC Cohorts
Mutations of the c-Met oncogene were first linked to changes in gene expressions in CRC at the genotypic level using the muTarget tool. Herein, high expression levels of MSANTD3 were associated with mutations on the c-Met gene compared to wild-type (WT) c-Met ( Figure 6A). Moreover, changes in the expression levels of the c-Met/GSK3β/MYC/CCND1 oncogenes to mutations of the top genes at the target level were also analyzed. Interestingly, we found that BRAF, NOTCH2, and ODF2 mutations were associated with the overexpression of c-Met, MYC, and CCND1 compared to the WT group ( Figure 6B-D). In a further analysis, we determined the alteration frequencies of c-Met/GSK3β/MYC/CCND1 gene signatures in CRC, COAD, and rectal adenocarcinoma (READ) tissues. Interestingly, the c-Met/GSK3β/MYC/CCND1 oncogenes were shown to be more mutated in CRC ( Figure 6E,F). We continued to analyze the genetic alterations from sample subtypes and found that c-Met/GSK3β/MYC/CCND1 were more altered in COAD chromosomal instability (CIN) and in microsatellite instability (MSI) and had higher numbers of mutations and amplifications in CRC tissues as shown in the bar graph ( Figure 6F). We also identified all the altered and unaltered genes associated with c-Met/GSK3β/MYC/CCND1 signatures ( Figure 6G) and further determined the distributions of c-Met/GSK3β/MYC/CCND1 mutations in CRC across the protein domains ( Figure 6H-J).
associated with the overexpression of c-Met, MYC, and CCND1 compared to the WT group ( Figure 6B-D). In a further analysis, we determined the alteration frequencies of c-Met/GSK3β/MYC/CCND1 gene signatures in CRC, COAD, and rectal adenocarcinoma (READ) tissues. Interestingly, the c-Met/GSK3β/MYC/CCND1 oncogenes were shown to be more mutated in CRC ( Figure 6E,F). We continued to analyze the genetic alterations from sample subtypes and found that c-Met/GSK3β/MYC/CCND1 were more altered in COAD chromosomal instability (CIN) and in microsatellite instability (MSI) and had higher numbers of mutations and amplifications in CRC tissues as shown in the bar graph ( Figure 6F). We also identified all the altered and unaltered genes associated with c-Met/GSK3β/MYC/CCND1 signatures ( Figure 6G) and further determined the distributions of c-Met/GSK3β/MYC/CCND1 mutations in CRC across the protein domains ( Figure 6H-J).

c-Met, MYC, and CCND1 Are Potential Drug Targets of NSC772864
We explored the DTP-COMPARE web server, and the analytical results show that c-Met, MYC, and CCND1 are potential targets of NSC772864. The predictions were based on an ARRAYCGH_GRAY analysis, and we further compared the compound's similarities to NCI synthetic compounds and standard agents as shown in Table 1. In a further analysis, we used the independent tool, PharmMapper, which identifies potential drug targets using a pharmacophore-matching approach. The analysis from the PharmMapper run was sorted according to the fit score. The results reveal the top 300 target genes of NSC772864, which were ranked by normalized fit scores ranging from 0.4875 to 0.9936. Interestingly, c-Met and CCND1 were identified among the 300 target genes for NSC772864. Accordingly, the c-Met/CCND1/cMYC-NSC772864 complex exhibited three hydrophobic interactions for c-Met and CCND1 and 15 hydrophobic interactions for c-MYC, with normalized fit scores of 0.6849, 0.6942, and 0.1667 ( Figure 8A-C), respectively. We explored the DTP-COMPARE web server, and the analytical results show that c-Met, MYC, and CCND1 are potential targets of NSC772864. The predictions were based on an ARRAYCGH_GRAY analysis, and we further compared the compound's similarities to NCI synthetic compounds and standard agents as shown in Table 1. In a further analysis, we used the independent tool, PharmMapper, which identifies potential drug targets using a pharmacophore-matching approach. The analysis from the PharmMapper run was sorted according to the fit score. The results reveal the top 300 target genes of NSC772864, which were ranked by normalized fit scores ranging from 0.4875 to 0.9936. Interestingly, c-Met and CCND1 were identified among the 300 target genes for NSC772864. Accordingly, the c-Met/CCND1/cMYC-NSC772864 complex exhibited three hydrophobic interactions for c-Met and CCND1 and 15 hydrophobic interactions for c-MYC, with normalized fit scores of 0.6849, 0.6942, and 0.1667 ( Figure  8A-C), respectively.

Drug Sensitivity Analysis for the c-Met/GSK3β/MYC/CCND1 Oncogenes in Colon Tumors
We explored the GSCA to determine the drug sensitivity profile data of c-Met/GSK3β/ MYC/CCND1 against different cancer cell lines. Herein, correlations of c-Met/GSK3β/MYC/ CCND1 gene expressions and drug sensitivity were analyzed based on values of the area under the dose-response curve (AUC) for drugs and gene expression profiles for all cell lines. The expression of each gene in the gene set was analyzed using a Spearman correlation analysis of the drug sensitivity (IC50) of the small molecule against various cell lines in the Cancer Therapeutics Response Portal (CTRP) and Genomics of Drug Sensitivity (GDSC) databases. The correlation coefficients indicated that increased gene expression levels were resistant to the drug. Herein, high messenger (m) RNA expression levels of c-Met and CCND1 (as indicated by orange circles) were associated with drug resistance (Figure 9).

Drug Sensitivity Analysis for the c-Met/GSK3β/MYC/CCND1 Oncogenes in Colon Tumors
We explored the GSCA to determine the drug sensitivity profile data of c-Met/GSK3β/MYC/CCND1 against different cancer cell lines. Herein, correlations of c-Met/GSK3β/MYC/CCND1 gene expressions and drug sensitivity were analyzed based on values of the area under the dose-response curve (AUC) for drugs and gene expression profiles for all cell lines. The expression of each gene in the gene set was analyzed using a Spearman correlation analysis of the drug sensitivity (IC50) of the small molecule against various cell lines in the Cancer Therapeutics Response Portal (CTRP) and Genomics of Drug Sensitivity (GDSC) databases. The correlation coefficients indicated that increased gene expression levels were resistant to the drug. Herein, high messenger (m) RNA expression levels of c-Met and CCND1 (as indicated by orange circles) were associated with drug resistance (Figure 9).

Evaluation of Drug Likeness, ADME Properties, and Friendliness of NSC772864
To evaluate the drug-likeness, PKs, medicinal chemistry, and ADME and toxicity (ADc-Met) properties of NSC772864, we used the swissADME online platform. Accordingly, we analyzed the drug-likeness properties of the compound according to Lipinski (rule of five), Egan, Ghose, and Veber standards. Moreover, we also showed associations of PKs with physiochemical properties. In a further analysis, we used the BBB penetration properties, using the brain or intestinal estimated permeation (BOILED-Egg) model [82,103], to predict the GIA and BBB penetration properties of NSC772864. The Abbot bioavailability score was determined based on the probability of the compound having at least 10% oral bioavailability in rats or measurable Caco-2 permeability [83] (Figure 10A,B). Drug target prediction software showed that NSC772864 has different targetable proteins, most of which were family A G protein-coupled receptors followed by kinases ( Figure 10C).  Positive Spearman correlation coefficients (orange circles) indicate that increased gene expression levels were resistant to the drug, compared to negative correlations shown in blue, which indicate sensitivity to the drug.

Evaluation of Drug Likeness, ADME Properties, and Friendliness of NSC772864
To evaluate the drug-likeness, PKs, medicinal chemistry, and ADME and toxicity (ADc-Met) properties of NSC772864, we used the swissADME online platform. Accordingly, we analyzed the drug-likeness properties of the compound according to Lipinski (rule of five), Egan, Ghose, and Veber standards. Moreover, we also showed associations of PKs with physiochemical properties. In a further analysis, we used the BBB penetration properties, using the brain or intestinal estimated permeation (BOILED-Egg) model [82,103], to predict the GIA and BBB penetration properties of NSC772864. The Abbot bioavailability score was determined based on the probability of the compound having at least 10% oral bioavailability in rats or measurable Caco-2 permeability [83] ( Figure 10A,B). Drug target prediction software showed that NSC772864 has different targetable proteins, most of which were family A G protein-coupled receptors followed by kinases ( Figure 10C).   (Table 2). Results are shown of the physiochemical properties, pharmacokinetics, drug-likeness, and medical chemistry.

In Vitro Anticancer Screening of SJ3 against NC1-60 CRC Cells
The NSC772864 compound was submitted to the NCI DTP for screening for potential in vitro anticancer activities against a panel of NCI CRC cell lines. NSC772864 was first tested at an initial dose of 10 µM. The compound exhibited antiproliferative and cytotoxic activities against CRC cell lines as indicated by the percentage growth altered by treatment. Accordingly, the growth inhibition percentage revealed antiproliferative effects against HCC-2988 cells of 88.65% and cytotoxic activities, showing KM12 to be more sensitive with -53.42% growth inhibition followed by HCT15 (−50 38%), COLO205 (−45.69%), SW620 (−43.72%), HCT116 (−36.96%), and HT29 cells (−9.75%) ( Figure 11A-I). Because NSC772864 showed potential antiproliferative and cytotoxic activities at an initial dose of 10 µM, the compound was further tested in a dose-dependent manner. Interestingly, the results show potential antiproliferative effects in a dose-dependent manner ( Figure 11J). We further investigated the in vitro GI50. The results ranged from 0. 16

Molecular Docking Analysis
To investigate the potential inhibitory effects of NSC772864, we conducted a molecular docking analysis. The results obtained from ligand-receptor complexes show unique different binding energies of NSC772864 with c-Met (ΔG = −8.0 kcal/mol, Figure   Figure 11. NSC772864 demonstrated anticancer effects in NC-I60 human colon cancer cell lines.

Molecular Docking Analysis
To investigate the potential inhibitory effects of NSC772864, we conducted a molecular docking analysis. The results obtained from ligand-receptor complexes show unique different binding energies of NSC772864 with c-Met (∆G = −8.0 kcal/mol, Figure 12A,B), GSK3β (=−8.6 kcal/mol, Figure 13A,B), MYC (∆G = −9.1 kcal/mol, Figure 14A,B), and CCND1 (∆G = −8.0 kcal/mol, Figure 15A,B). These results were further compared to standard inhibitors of crizotinib for c-Met ( Figure 12C,D), AZD1080 for GSK3β ( Figure 13C,D), alobresib for MYC. (Figure 14C,D), and trilacoclib for CCND1 ( Figure 15C,D), with respective binding energies of ∆G = −8.1, −8.4, −7.6, and −7.4 kcal/mol. Interestingly, NSC772864 showed higher binding energies compared to AZD1080, alobresib, and trilacoclib; however, crizotinib was shown to possess a slightly higher binding energy compared to NSC772864. In a further analysis, we used Pymol and Discovery Studio for the analysis and visualization of the results. The NSC772864/c-Met complex showed interactions of conventional hydrogen bonds (H-bonds) with ASP1222 (2.20 Å), GLU1127 (1.85 Å), and LYS1110 (2.68 Å) compared to the crizotinib/c-Met complex, which showed Hbonds with ASP1222 (2.20 Å), GLU1258 (2.19 Å), ASP1204 (1.96 Å), and ARG1227 (2.60 Å). The NSC772864/GSK3β complex displayed H-bond interactions with ASN64 (2.56 Å), compared to the AZD1080/GSK3β complex, which displayed H-bond interactions with PHE67(2.30Å) and GLN185 (2.70 Å). The NSC772864/myc complex displayed H-bond interactions with SER221 (2.48 Å), compared to the alobresib/myc complex, which had no H-bond interactions. Moreover, the NSC772864/ccnd1 complex exhibited an H-bond interaction with ARG26 (2.88 Å), compared to trilacoclib/ccnd1, which showed H-bond interactions with HIS163 (2.29 Å) and ALA133 (2.04 Å). The interactions were sustained by van der Waals forces, carbon-hydrogen bonds, π bonds, and alkyl bonds as shown in (Tables 2-4).  Å). The interactions were sustained by van der Waals forces, carbon-hydrogen bonds, π bonds, and alkyl bonds as shown in (Tables 2-4).  Table 3 gives binding energies of ligand-receptor interactions, including different types of interactions and the amino acid residues involved.   Table 3 gives binding energies of ligand-receptor interactions, including different types of interactions and the amino acid residues involved.   Table 4 gives binding energies of ligand-receptor interactions, including different types of interactions and the amino acid residues involved.   Table 4 gives binding energies of ligand-receptor interactions, including different types of interactions and the amino acid residues involved.   Table 5 gi binding energies of ligand-receptor interactions, including different types of interactions and amino acid residues involved.   Table 5 gives binding energies of ligand-receptor interactions, including different types of interactions and the amino acid residues involved.   Table 6 gives the binding energies of ligand-receptor interactions, including different types of interactions and the amino acid residues involved.   Table 6 gives the binding energies of ligand-receptor interactions, including different types of interactions and the amino acid residues involved.

Discussion
In recent years, the survival rates of CRC patients have significantly improved thanks to advancements in novel therapeutics, including surgical resection, chemoradiotherapy, and targeted therapy [45,[104][105][106]. The 5-year survival rate at the initial cancer stage is almost 90% [107]; however, many patients still experience resistance to these drugs and develop distant metastasis [108]. Thus, there is an urgent need to identify novel and specific biomarkers, which will assist in the early detection of CRC and improve strategies for drug development [9,15,109]. c-Met tyrosine kinase was reported to be a prognostic marker in primary and metastatic cancer [110]. Studies showed that c-Met plays a significant role in cancer progression in different cancers [13,14,20,111,112]; moreover, dysregulation of the c-Met pathway was demonstrated in CRC, although still little is known about its genetic mutations in CRC [110,111,113]. In addition, others showed that c-Met is expressed in approximately 60% of CRC patients, and its upregulation is associated with disease progression [113].
Accumulating studies have demonstrated that c-Met is a target gene of WNT/βcatenin signaling [114], which significantly promotes the deregulation of the protoMYC oncogene in CRC. MYC has been extensively investigated as a predominant cancer-causing gene and is associated with the cell cycle, differentiation, and apoptosis [115]. Several studies reported the overexpression of c-MYC in almost 10% of CRC patients [38,41,42]. Those findings suggest a potential treatment approach targeting c-MYC signaling. However, acquiring targeted therapies for c-MYC remains a huge challenge because MYC is located in the nucleus [46]. Previously, there were satisfactory breakthroughs in describing critical cell cycle signaling pathways in cancer initiation and progression. CCND1 was one of the identified pathways involved [116]. Studies showed that the overexpression of c-MYC enhances the expression of CCND1 through beta-catenin activation [45]. CCND1 is a crucial checkpoint protein that regulates the transition of the G1 to S phase; moreover, CCND1 was shown to promote tumorigenesis, invasion, and resistance to apoptosis [117].
In CRC, CCND1 overexpression has yielded conflicting results, as some researchers demonstrated its role in promoting poor clinical outcomes, while others reported it as a good prognostic marker [117][118][119][120]. Those findings thus suggest crosstalk among c-Met/GSK3β/MYC/CCND1 in regulating the cell cycle in CRC. Herein, we conducted a bioinformatics simulation and identified the 25 top DEGs in CRC. Interestingly, we found the c-Met/GSK3β/MYC/CCND1 signaling pathway to be expressed among these top 25 genes (Figure 2). In a further analysis, we applied a PPI network (PIN) functional enrichment analysis. We discovered that c-Met/GSK3β/MYC/CCND1 were coexpressed in the same clustering network. When we further performed an enrichment analysis, we found that enriched GO BPs involved the transcription factor network, G1 phase, cyclin D, and mitosis G1-G1/S phases ( Figure 3). Recently, considerable attention has focused on small molecules as targeted therapies for cancer treatment [35]. In this study, we demonstrated the anticancer activities of our newly synthesized quercetin-derived compound, NSC772864, against panels of NCI human CRC cell lines. The compound exhibited cytotoxic activities against CRC cell lines as indicated by the percentage growth altered by treatment; the growth inhibition percentage revealed that HCC-2988 cells were more sensitive to the compound with growth inhibition of −88.65%, followed by KM12 (at −53.42%), HCT15 (at −50.38%), COLO205 (at −45.69%), SW620 (at −43.72%), HCT116 (at −36.96%), and HT29 cells (at −9.75%).
Moreover, we utilized the swissADME web platform and found that NSC772864 passed the drug-likeness, PK, medicinal chemistry, and ADMET criteria, as shown in Figure 5. We further used the DTP-COMPARE platform and PharmMapper software and predicted the c-Met, c-MYC, and CCND1 signatures as target genes of NSC772864 ( Figure 16). To validate these results, we evaluated protein-ligand interactions through a molecular docking analysis. The results obtained from the ligand-receptor complexes show unique, different binding energies of NSC772864 with c-Met (∆G = −8.0 kcal/mol), GSK3β (∆G = −9.1 kcal/mol), MYC (∆G = −9.1 kcal/mol), and CCND1 (∆G = −8.0 kcal/mol). These results were further compared to standard inhibitors of crizotinib for c-Met, AZD1080 for GSK3β, alobresib for MYC, and trilacoclib for CCND1, with respective binding energies of ∆G = −8.1, −8.6, −7.6, and −7.4 kcal/mol. These results suggest that NSC772864 is a potential and novel compound for treating c-Met/GSK3β/MYC/CCND1 in CRC. Collectively, these findings create an opportunity for a novel research approach for NSC772864 as a novel and potential antiCRC agent. , x FOR PEER REVIEW 22 of 28 Figure 16. Schematic diagram shows NSC772864 may directly target Met receptor-mediated cell cycle progression through WNT/β-catenin/GSK3β pathway and MYC transcription factor. c-Met expression enhances the Wnt/β-catenin signaling and impedes GSK3β from phosphorylating βcatenin in CRC; this subsequently promotes β-catenin translocation into the nucleus leading to cancer initiation; moreover, proto-oncogene, c-MYC, is associated with the regulation of CCND1, a key regulatory protein through the G1 to S phases of the cell cycle via binding to cyclin-dependent kinase 4 (CDK4) and CDK6 cell cycle, and play crucial roles in cell cycle progression. Therefore, blockade of c-Met/cMYC/CCND1 inhibits CRC cell progression, stemness, and metastasis.

Conclusions
We identified c-Met/GSK3β/MYC/CCND1 oncogenic signature as major players in CRC progression through the cell cycle. Our bioinformatics analysis demonstrated that these oncogenic signatures are overexpressed in CRC tissues compared to adjacent normal tissues, resulting in tumor progression, metastasis, poor prognoses, and resistance Figure 16. Schematic diagram shows NSC772864 may directly target Met receptor-mediated cell cycle progression through WNT/β-catenin/GSK3β pathway and MYC transcription factor. c-Met expression enhances the Wnt/β-catenin signaling and impedes GSK3β from phosphorylating βcatenin in CRC; this subsequently promotes β-catenin translocation into the nucleus leading to cancer initiation; moreover, proto-oncogene, c-MYC, is associated with the regulation of CCND1, a key regulatory protein through the G1 to S phases of the cell cycle via binding to cyclin-dependent kinase 4 (CDK4) and CDK6 cell cycle, and play crucial roles in cell cycle progression. Therefore, blockade of c-Met/cMYC/CCND1 inhibits CRC cell progression, stemness, and metastasis.

Conclusions
We identified c-Met/GSK3β/MYC/CCND1 oncogenic signature as major players in CRC progression through the cell cycle. Our bioinformatics analysis demonstrated that these oncogenic signatures are overexpressed in CRC tissues compared to adjacent normal tissues, resulting in tumor progression, metastasis, poor prognoses, and resistance to current chemotherapies and targeted therapies. We further demonstrated the anticancer activities of our newly synthesized quercetin-derived small molecule, NSC772864, against NCI human CRC cell lines and its dose-dependent cytotoxic preference for CRC. Our target prediction analysis showed that c-Met/GSK3β/MYC/CCND1 signatures are target genes for NSC772864, and these findings were validated thought a molecular docking analysis, which showed remarkable binding energies of NSC772864 in a complex with c-Met, GSK3β, c-MYC, and CCND1. These findings create an opportunity for a novel research approach for NSC772864 as a novel and potential antiCRC agent. Further in vitro and in vivo investigations to validate the antiproliferative and cytotoxic activities of NSC772864 are ongoing. Data Availability Statement: Data supporting the results of this manuscript will be made available without undue reservation.

Conflicts of Interest:
The authors declare no conflict of interest.