mTOR/EGFR/iNOS/MAP2K1/FGFR/TGFB1 Are Druggable Candidates for N-(2,4-Difluorophenyl)-2′,4′-Difluoro-4-Hydroxybiphenyl-3-Carboxamide (NSC765598), With Consequent Anticancer Implications

Background The application of computational and multi-omics approaches has aided our understanding of carcinogenesis and the development of therapeutic strategies. NSC765598 is a novel small molecule derivative of salicylanilide. This study aims to investigate the ligand-protein interactions of NSC765598 with its potential targets and to evaluate its anticancer activities in vitro. Methods We used multi-computational tools and clinical databases, respectively, to identify the potential drug target for NSC765598 and analyze the genetic profile and prognostic relevance of the targets in multiple cancers. We evaluated the in vitro anticancer activities against the National Cancer Institute 60 (NCI60) human tumor cell lines and used molecular docking to study the ligand-protein interactions. Finally, we used the DTP-COMPARE algorithm to compare the NSC765598 anticancer fingerprints with NCI standard agents. Results We identified mammalian target of rapamycin (mTOR)/epidermal growth factor receptor (EGFR)/inducible nitric oxide synthase (iNOS)/mitogen-activated protein 2 kinase 1 (MAP2K1)/fibroblast growth factor receptor (FGFR)/transforming growth factor-β1 (TGFB1) as potential targets for NSC765598. The targets were enriched in cancer-associated pathways, were overexpressed and were of prognostic relevance in multiple cancers. Among the identified targets, genetic alterations occurred most frequently in EGFR (7%), particularly in glioblastoma, esophageal squamous cell cancer, head and neck squamous cell cancer, and non–small-cell lung cancer, and were associated with poor prognoses and survival of patients, while other targets were less frequently altered. NSC765598 displayed selective antiproliferative and cytotoxic preferences for NSCLC (50% growth inhibition (GI50) = 1.12–3.95 µM; total growth inhibition (TGI) = 3.72–16.60 μM), leukemia (GI50 = 1.20–3.10 µM; TGI = 3.90–12.70 μM), melanoma (GI50 = 1.45–3.59 µM), and renal cancer (GI50 = 1.38–3.40 µM; TGI = 4.84–13.70 μM) cell lines, while panels of colon, breast, ovarian, prostate, and central nervous system (CNS) cancer cell lines were less sensitive to NSC765598. Interestingly, NSC765598 docked well into the binding cavity of the targets by conventional H-bonds, van der Waal forces, and a variety of π-interactions, with higher preferences for EGFR (ΔG = −11.0 kcal/mol), NOS2 (ΔG = −11.0 kcal/mol), and mTOR (ΔG = −8.8 kcal/mol). NSC765598 shares similar anti-cancer fingerprints with NCI standard agents displayed acceptable physicochemical values and met the criteria of drug-likeness. Conclusion NSC765598 displayed significant anticancer and potential multi-target properties, thus serve as a novel candidate worthy of further preclinical studies.


INTRODUCTION
Currently, cancer is one of the leading causes of global morbidity and the second leading cause of mortality worldwide (1), accounting for an estimated 19.3 million new cases and 9.9 million deaths in 2020 (2). One in five men or women develops cancer, while death occurs in 1 out of 8 men or 1 out of 11 women. However, due to increasing rates of predisposing factors, such as smoking, a sedentary lifestyle, being overweight, and increasing pollution associated with urbanization and industrial development, cancer is predicted to top the rank of leading causes of global mortality and be the most important independent dictator of a poor life expectancy in the 21st century (3). In 2020, female breast cancer has surpassed lung cancer as the most commonly diagnosed cancer accounting for 11.7% (2.3 million new cases) (2), and closely followed by lung (11.7%), colorectal (10.0%), prostate (7.3%), and stomach (7.7%) cancers, whereas colorectal (9.4% deaths), liver (8.3% death), stomach (7.7% deaths), and breast (6.9% deaths) cancers followed lung cancer (with an estimated 1.8 million deaths (18%) in terms of mortality rates (2), with higher incidences in males than in females (4). Among women, breast cancer is the most frequently diagnosed and a prominent cause of cancer deaths, followed by colorectal, lung, and cervical cancers (3). Unfortunately, the global cancer burden is expected to be 28.4 million cases in 2040, a 47% rise from 2020 (2). This projection is solely due to the growth and aging of the population and an increasing prevalence of risk factors.
The most common strategy for managing cancer is conventional therapies involving surgery and chemo-and radiotherapies. However, despite the early effectiveness demonstrated by these oncological strategies (5,6), therapeutic resistance and toxicity, among other factors, have limited the success rates and long-term survival of cancer patients. Hence, research is needed on novel therapeutic options, which can offer better and more-specific multi-target antitumor activities (7), with minimal side effects to ultimately enhance the survival of cancer patients. Without a doubt, multitarget small molecules with anticancer activities represent alternative strategies with promising features for the realization of novel drugs that could target multiple aberrant signaling pathways (8,9), offering fewer chances for drug resistance and thus better prognoses of cancer patients.
Several potential molecular targets exist for exploring the dysregulation of signaling pathways of cancer cells (10)(11)(12). Growth factors, including epidermal growth factor (EGF), fibroblast growth factor (FGF), transforming growth factor (TGF), and vascular endothelial growth factor (VEGF), are compact molecules that play important roles in regulating cellular communication, growth and differentiation, proliferation, survival, migration, and metastasis of cancer cells (13,14). Upon receptor binding, EGF and TGF activate downstream mitogen-activated protein kinase (MAPK) and AKT-mammalian target of rapamycin (mTOR) pathways (15), which regulate apoptosis and cell proliferation (16), while inducible nitric oxide synthase (iNOS) modulates cellular bioenergetics pathways in cancer (17). Therefore, these signaling pathways are important for cancer initiation, progression, and metastasis and thus act as attractive targets for anticancer therapy exploration. Mapping these targets with corresponding drugs, therefore, has become a relevant prerequisite for drug discovery.
Over the years, rational drug design (RDD) based on computational biology and bioinformatics has not only changed the way drugs are designed but also speed up drug discovery and development processes while reducing costs and the wasting of resources (18). Computational and bioinformatics studies of cancer involve a variety of methods including mining and analysis of clinical data to give insights into the prognostic relevance of gene signatures, analyze geno-phenotypic responses to clinical drugs, and identify novel compounds with their potential targets before in vitro and in vivo validation studies. One such method is molecular modeling and docking studies of receptor-ligand interactions. Molecular docking is frequently employed to predict binding orientations of small molecules with their target proteins in order to calculate the affinities and activities of the small-molecule drug candidates (18,19).
In our previous study, a salicylanilide analogue, NDMC101 (PubChem CID: 60202556), was synthesized, and we observed significant cytotoxic effects via downregulation of proliferative and inflammatory markers (20). A series of modifications of the salicylanilide core scaffold of the lead molecule were synthesized, and their preclinical anticancer activities against a variety of cancers were reported (9,(21)(22)(23). As a continuing effort to find and screen more active derivatives of NDMC101, in the present study, we reported a new derivative, NSC765598, for anticancer activities and identified mTOR/EGR receptor (EGFR)/iNOS/ MAP2K/TGF-b1 (TGFB1)/fibroblast growth factor receptor-1 (FGFR1) as its potential targets.

Clinical Data Mining of Databases
Differential gene expression profiles of the EGFR, iNOS, mTOR, FGF receptor (FGFR), TGF receptor (TGFR), and MAP2K1 across The Cancer Genome Atlas (TCGA) databases were analyzed using the Tumor IMmune Estimation Resource (TIMER2.0) resource (http://timer.cistrome.org/) (28). The prognostic relevance of the genes in 9736 tumor samples across 33 TCGA cancer types was analyzed using the Gene Expression Profiling Interactive Analysis (GEPIA) (http://gepia. cancer-pku.cn/) (29). We collected the RNA expression profile of the genes and then set the median expression as the expression threshold to split the patient samples into high-expression and low-expression groups, and used the Kaplan-Meier survival plot to assess the overall survival (OS) with the hazard ratio (HR), a 95% confidence interval (CI), and a log-rank test p-value. We also explored the cancer genomic dataset using the cBioPortal tool (http://www.cbioportal.org/) to analyze genomic alterations and survival and perform group comparisons of EGFR, iNOS, mTOR, FGFR, TGFR, and MAP2K1 in 10,953 cancer patients (10,967 samples) from different cancer types (30,31).  (32) and Enrich (https://maayanlab. cloud/Enrichr/enrich#) (33,34). KEGG enrichment criteria were set to p < 10 −5 , gene count of ≥ 3, and strength of ≥ 2.

Receptor-Ligand Interaction Study
The Three dose-response parameters including the drug concentration causing total growth inhibition (TGI), drug concentration causing a 50% reduction in the net protein increase in control cells during the drug incubation (GI 50 ), and drug concentration causing 50% cell lethality (LC 50 ), indicating a net loss of cells following treatment) (42) were computed. GI 50 is calculated from [(Ti-Tz)/(C-Tz)] × 100 = 50, TGI is calculated from Ti = Tz, while the LC 50 is calculated from [(Ti-Tz)/Tz] × 100 = −50. Values are calculated for each of these three parameters if the level of activity is reached; however, if the effect is not reached or is exceeded, the value for that parameter is expressed as greater or less than the maximum or minimum concentration tested.

DTP-COMPARE Analysis of NSC765598 Anticancer Fingerprints
The activity patterns (fingerprints) of NSC765598 were correlated to NCI synthetic compounds and standard agents using the DTP-COMPARE algorithms (43). The NSC numerical IDs were used as "seed" while GI 50 , TGI, and LC 50 were set as the endpoints (8).

Data Analysis and Visualization
Data were visualized using GraphPad Prism 8 software. Overall survival of the cohorts was visualized in Kaplan-Meier (K-M) survival plots. The statistical significance of differentially expressed genes was evaluated using the Wilcoxon test. *p<0.05; **p<0.01; and ***p<0.001 indicate statistical significance. Growth inhibition was calculated relative to cells without drug treatment and the time-zero control. The growth inhibition by NSC765598 in the single-dose assay was obtained by subtracting the positive value on the plot from 100, i.e., a value of 60 would indicate 40% growth inhibition.

RESULTS
Drug-Likeness, Physicochemical, and ADMET Properties of NSC765598 NSC765598 was synthesized as a white powder purity of > 95%. Our evaluation of the drug-properties revealed that NSC765598 exhibited acceptable physicochemical and ADMET properties, and satisfied Lipinski's rule of five for drug-likeness. It has good synthetic accessibility (2.27), a high bioavailability score (0.607-0.87), and low toxicity (LD 50 = 880.766 mg/kg). Interestingly, NSC765598 exhibited a high human intestinal absorption capability and was permeant to the blood-brain barrier (Table 1) Table 1). Furthermore, NSC765598 target predictions using SwissTarget algorithm partitioned the top 20 targets into three classes: enzymes (20%), kinases (40%), and family A G protein-coupled receptors (40%) ( Figure 1B). Coherent with the PASS target predictions, SwissTarget also identified EGFR, TGFB1, FGFR1, iNOS, VEGFR2, MAPK, and serine/threonine-protein kinase mTOR as top druggable targets of NSC765598 (Supplementary Table 2). Furthermore, we also used PharmMapper Server, an integrated pharmacophore matching platform with statistical methods to identify potential targets. Interestingly out of 1627 drug targets (459 of which were human protein targets) on PharmMapper, EGFR, TGF-bR1, FGFR1 and −2, VEGFR2, iNOS, MAPK, HGFR, insulin-like growth factor 1 receptor (IGF1R), B-Raf proto-oncogene serine/ threonine-protein kinase, serine/threonine-protein kinase Chk1, and cyclin-dependent kinase 5 (CDK5) activator 1 were predicted to be among the top 20 targets for NSC765598. The top 20 predicted targets exhibited three or more total protein-ligand interactions, two or more hydrophobic interactions with a normalized fit score range of 0.7082 to 0.9998, and a z score range of 0.1056 to 0.641082 (Supplementary Table 3). Collectively, six genes, viz., iNOS, TGFB1, VEGFR2, FGFR, EGFR, and MAPK2 were predicted by the three algorithms ( Figures 2A, B), while two genes, viz., signal transducer and activator of transcription (STAT) and serine/threonine-protein kinase (mTOR) were predicted to be NSC765598 top ranked target by 2 algorithms.
Our correlation analysis revealed that EGFR expression was strongly correlated with expressions of mTOR (r= 0.262-0.496) and FGFR1 (r= 0.203-0.348) in liver, lung, kidney, glioblastoma, colon, head and neck, and breast cancers. MAP2K1 expression was strongly correlated with expressions of EGFR (r= 0.252-0.62) in liver, lung, kidney, colon, head and neck, and breast cancers but negatively correlated in glioblastoma (r = −0.062). EGFR expression was strongly correlated with expressions of NOS2 (r= 0.287-0.388) in liver cancer, kidney cancer and, breast cancer, poorly correlated in head and neck cancer (r=0.082) but show no association (r = −0.001) in colon and lung cancer ( Figure 5).

Molecular Docking of NSC765598 With EGFR
Molecular docking of NSC765598 with EGFR revealed more robust interactions and higher binding affinity (−11 kcal/mol) to EGFR than do gefitinib (−6.8 kcal/mol) a standard EGFR inhibitor. The NSC765598-EGFR complex was bonded by strong H-bond interactions with GLU762 and CYS797, halogen bond interactions with ASN842, ARG841, and ASP855, and multiple p-interactions, including p-p interaction with PHE723, p-alkyl interaction with LEU844, and p-sulfur interactions with MET790 and CYS797 ( Figure 7C). In terms of interaction distances between the ligand atoms and receptor atoms, the hydrogen bond interactions of GLU762 (2.42 Å) and CYS797 (2.29 Å) were the shortest, while p-interactions of LEU844 (5.49 Å), CYS797 (3.63 Å), and MET790 (5.43 Å) had longer distance interactions within the NSC765598-EGFR complex. NSC765598 also forms 5 hydrophobic contacts with PHE723 and LEU844 residues of EGFR. Stabilization of the NSC765598-EGFR complex was also supported by Van der Waal forces between the ligand and ARG858, LYS745, LEU788, and THR854 in receptor binding pockets ( Figure 7C, Table 3).

Molecular Docking of NSC765598 With FGFR1
NSC765598 docked well with the binding pocket of FGFR1 with a binding affinity of −7.3 kcal/mol stronger than the interactions between FGFR and Erdafitinib (−5.7 Kcal/mol). The NSC765598-FGFR1 complex was stabilized by two H-bonds with SER119 (2.13 Å) and VAL121 (3.33 Å) as opposed to only single H-bond interaction (ARG134) between FGFR_Erdafitinib. In addition, halogen interactions (PRO120, 2.54 Ăand PRO149, 3.05 Ă), p-p stacking with the aromatic ring of PHE129, and Amide-Pi stacking with SER119 were also observed between the NSC765598_FGFR1 complex. NSC765598-FGFR1 complex was also supported by several Van der Waal forces created around the backbone of the ligand with respective amino acid residues (LYS81, GLU117, ARG134, PHE136, PHE151, PHE150, and SER127) of the receptor binding pocket ( Figure 8A, Table 4).

Molecular Docking of NSC765598 With iNOS
Molecular docking of NSC765598 with iNOS revealed a binding affinity of −11.0 kcal/mol. NSC765598 interacts with iNOS by three H-bonds with TYR483, GLY196, and GLY363, and halogen interactions with GLN257, PRO344, and THR184. Binding interaction proximities in the NSC765598-iNOS complex  Å). The NSC765598-iNOS complex was also supported by p-interactions; p-p stacking with aromatic rings of TRP188 and PHE363, and p-alkyl interactions with CYS194, LEU203, and PRO344. The interaction was also stabilized by Van der Waal forces created on the backbone with the following amino acids: ILE238, SER236, GLN199, TRP366, MET368, TYR367, ALA345, VAL346, and ASN364 ( Figure 8B, Table 4). Furthermore, NSC765598 interaction with iNOS is more robust and with higher affinity than the interactions between iNOS and N-Iminoethyl-L-lysine dihydrochloride (−5.3 Kcal/mol), a standard iNOS inhibitor (Table 4). In addition, NSC765598 forms 4 hydrophobic contact with TRP188, PRO344, and PHE363 residue of iNOS as compared to 1 hydrophobic contact in the iNOS_N-Iminoethyl-Llysine dihydrochloride complex

Anti-Proliferative Activities of NSC765598 Against NCI60 Panels of Human Tumor Cell Lines
The percent growth inhibition (GI) caused by single-dose testing revealed that NSC765598 inhibited the growth of all of the NCI60 cell line panels of breast, prostate, renal, ovarian, colon, melanoma, CNS, leukemia, and non-small cell lung cancers. Single-dose (10 mM) treatment with NSC765598 exhibited more than 50% growth inhibition of all the NCI60 cell line panels except for two colon cancer cell lines (HT29 and COLO 205) which were completely insensitive to NSC765598 treatment ( Figure 9). However, no cytotoxic activities were detected against the ovarian, colon, or CNS cancer cell lines (Figure 9). The primary single-dose results clearly  GLN312  VAL182  ARG858  GLY719  PRO21146,  ILE1939  LYS179  ALA181  LYS745  LYS745,  ALA2226,  GLU2196  ILE146  TYR174  LEU788  PHE723,  GLU2196,  MET2199  ALA181  ILE146  THR854  MET790,  MET2199,  GLN2200  LEU183  ARG133  ARG841,  PRO1940  GLY2203  HIS96  PHE138  ASN842,  GLY2203  THR2207  LYS149  THR854  LEU2204  ASN2147  GLN312  LEU792,  LEU1900  ARG141  GLY796,  ASN1899 PRO794 p -sulfur: p-electron cloud between the Aromatic rings of ligands and lone pair of electron cloud of sulfur atom in the receptors; p -p stacked: p-electron cloud between the Aromatic rings, p -p T-shaped: T shaped p-electron cloud between the Aromatic rings, p -alkyl; p-electron cloud between the Aromatic ring of ligand and alkyl group of ligand.
indicated the anticancer activities of NSC765598 against different kinds of human cancer cell lines, and thus it is worthy of further evaluation for dose-dependent activities.

NSC765598 Exhibited Selective Cytotoxic Preferences for NSCLC, Leukemia, Melanoma, and Renal Cancer Cell Lines
In the five-dose assay screening, NSC765598 demonstrated selective dose-dependent cytotoxic effects against the panel of NCI60 human tumor cell lines ( Figure 10). The GI 50 values of NSC765598 (concentration of NSC765598 that causes 50% inhibition of cell growth) against individual cells of the NCI60 panels of human cell lines tested were lower than 5.0 µM, except for COLO 205, a colon cancer cell line with a GI 50 of 16.0 µM ( Figure 11). However, among the nine types of cancer involved in this assay, panels of NSCLC cell lines were the most responsive to growth inhibition by NSC765598 treatment (GI 50 = 1.12-3. . However, contrary to the primary one-dose cell growth percent inhibition assay, the colon cancer cell line, COLO 205 was found to be the most sensitive to five-dose testing of NSC765598 than the other colon cancer cell lines, displaying total growth inhibition and 50% cytotoxic response at NSC765598 concentrations of 35.80 and 80.30 mM, respectively ( Figure 11, Table 5).

NSC765598 Share a Similar Antitumor Fingerprint With NCI Synthetic Compounds and Standard Agents
DTP-COMPARE analysis indicated that NSC765598 share similar antitumor fingerprints with a number of NCI standard agents. The p-value and cell counts of the top 15 NCI correlated standard agents are shown in Table 6. All NCI synthetic compounds with NSC765598 similar antitumor fingerprints p-sulfur, p-electron cloud between the Aromatic rings of ligands and lone pair of electron cloud of sulfur atom in the receptors; p-p stacked: p-electron cloud between the Aromatic rings, pp T-shaped: T shaped p-electron cloud between the Aromatic rings, p-alkyl; p-electron cloud between the aromatic ring of ligand and alkyl group of the receptor, N-I-LD, N-Iminoethyl-Llysine dihydrochloride.
are small molecules (MW; 277.3-461.9 g/mol). The identity, p-value, cell counts, and molecular weight of the top 15 NCI synthetic compounds are also shown in Table 6.

DISCUSSION
The application of multi-omics approaches has aided our understanding of carcinogenesis and the development of therapeutic strategies (44). Identifying drug targets for small molecules and their molecular mechanisms play important roles in drug design, discovery, and development (45). Drug targets can be identified via genetic interactions, biochemical methods, and computational studies (46). In this study, our collective computational approach identified EGFR/mTOR/NOS2/ TGFB1/FGFR1/MAP2K1 as potential targets for NSC765598. The PPI network revealed that these gene targets interact strongly with other oncogenic proteins and were mainly enriched in pathways and processes associated with multiple cancers. In accordance with our observations, previous studies had identified these gene signatures as cancer-associated genes. For instance, Murugan et al. (47) reported that mTOR promotes cancer growth, drug resistance, and metastasis, while McCubrey et al. (48) reported that the EGFR/mTOR signaling pathways play prominent roles in malignant transformation, therapeutic resistance, cancer stemness, metastasis, and inhibition of apoptosis. MAP2K was implicated in hypoxia-mediated angiogenesis, cancer progression, metastasis, and drug resistance (49). Increasing evidence also implicated EGFR signaling pathways in promoting cancer growth and metastasis (50,51). Taking together, these results strongly suggest the oncogenic roles of EGFR/mTOR/NOS2/TGFB1/FGFR1/ MAP2K1 signatures in multiple cancers. Hence, we explored public databases to investigate the full prognostic relevance of these target genes in multiple cancers. Comparative gene expression profiling of cancer and adjacent normal tissues facilitates an understanding of disease etiologies and enhances the design of appropriate therapeutic interventions (16). Our differential expression analysis revealed that mRNAs of EGFR, mTOR, NOS2, TGFB1, FGFR1, and MAP2K1 were overexpressed in multiple cancer types compared to corresponding adjacent healthy cohorts, while the survival analysis revealed that cohorts with higher expressions of EGFR, MAP2K1, and NOS2 exhibited shorter OS than cohorts with low expression profiles. Collectively, these data suggested that EGFR/mTOR/NOS2/TGFB1/FGFR1/ MAP2K1 play important roles in the progression of tumors and serve as reliable biomarkers of poor prognosis Using the cBioPortal, we also studied frequencies of genetic alterations of these signatures and their prognostic values in cancer patients. We found that genetic alterations in EGFR occurred in 7% of cancer cohorts, most frequently in GBM, ESCC, HNSCC, and NSCLC, and were associated with poor prognoses and survival of patients. Dong et al. (52) evaluated the alteration status of the top three genes identified as prognosticators of renal cell carcinoma and respectively found only 2%, 2%, and 1.7% genetic alterations in the FN1, COL1A2, and COL3A1 genes. Chen et al. (53) studied genetic alterations of 20 genes and found alteration frequencies ranging 0.8% to 3% for 19 individual genes, while only one gene (NBN) had an alteration frequency of 9%. Therefore, percentages of patients with genetic alterations in EGFR (7%), FGFR (6%), mTOR (4%), and iNOS (3%) as reported in this study were on the high side and thus strengthen our earlier observations that these signatures are very important in cancer progression and can thus serve as attractive targets worthy of further exploration.
Nevertheless, it is becoming increasingly evident that the initiation and progression of cancer cannot be ascribed to a single genetic mutation (54). Therefore, we evaluated the cooccurrence of genetically altered EGFR with other genetic mutations in the cancer cohorts, and we found its co-occurrence with genetically altered RICTOR, UGGT1, ATRNL1, AHNAK2, RIMS2, DSP, STXBP5L, DUOX2, PKD1L1, and MYH3. Alterations in these genes were found to be positively enriched in patients with altered EGFR, while alterations in KRAS and IDH1 were only enriched in patients without genetically altered EGFR. Intriguingly, we also observed that the most common genetic alterations in the EGFR gene were amplifications and gene gains, supporting the role of EGFR as an oncoprotein. These findings are also coherent with our earlier observations on the differential expression of EGFR mRNA in cancer cohorts compared to healthy cohorts. However, contrary to our earlier observations on the poor survival rates of cohorts with high mRNA expressions of EGFR, mTOR, FGFR iNOS, TGFB1, and MAP2K1, only genetic alterations of EGFR, but not those of mTOR, iNOS, TGFB1, or MAP2K1, were associated with poor prognoses and survival of patients.
Lung cancer is the most commonly diagnosed (often diagnosed in advanced stages) and the leading cause of global FIGURE 10 | Dose-response curves of NCI60 human cancer cell lines to NSC765598 treatment. Growth percentage value of 100 on the y-axis represents the growth of untreated cells, the 0 value represents no net growth, while −100 represents complete death of cells. cancer mortality due to limited therapeutic success (3,55). It is therefore plausible to say that genetic alterations in EGFR are a major contributor to this dilemma. In line with our observations, Ding et al (56). reported that the EGFR pathway is the most altered pathway in lung adenocarcinomas (26%). Gower et al (57). identified that alterations in EGFR are major oncogenic drivers of acquired resistance and predict poor survival in NSCLC, while Saadeh et al (58). reported high correlations of EGFR mutations with poor therapeutic responses and survival of glioblastoma patients. Although, treatments with EGFR tyrosine kinase inhibitors (TKIs) showed initial therapeutic responses, acquired resistance to 1st-, 2nd-, and recently approved 3rdgeneration TKIs (osimertinib) owing to EGFR mutations has been well reported (59,60). Taken together, patients with high EGFR mRNA expression are more likely to harbor genetically altered EGFR and experience worse prognoses. This necessitates alternative treatment strategies; hence, we evaluated NSC765598 for that purpose.
In order to facilitate the identification of potent anticancer small molecules from synthetic libraries, the National Cancer Institute (NCI) established the NCI60 cell panel screen to examine the activities of small molecules on cancer cell viability (61). Hence, we explored this avenue to evaluate the anticancer properties of NSC765598 against the NCI60 panel of cell lines. Interestingly we found that NSC765598 exhibited antiproliferative effects on all panels of breast cancer, prostate cancer, renal cancer, ovarian cancer, melanoma, CNS cancer, leukemia, and NSCLC cells with GI 50 values of < 5 µM, but it showed lower anti-proliferative activities against colon cancer cell lines. Among the nine types of cancer involved in this assay, NSC765598 had some cytotoxic preference for NSCLC, leukemia, melanoma, and renal cancer cell lines. Interestingly, NSC765598 also exhibited activity against a drug-resistant cell line, NCI/ADR-RES (GI 50 = 2.54 µM; TGI = 8.00 mM). However, the least amounts of cell lethality were found against panels of colon, breast, ovarian, prostate, and CNS cancer cell lines (TGI > 20 µM in most cases and LC 50 >100 µM in almost all cases), suggesting that NSC765598 is not generally toxic to growing cell lines, but displays some degree of specificity to NSCLC, renal cancer, leukemia, and melanomas. The NCI60 cell lines used in this study have been well characterized for proteins, genes, microRNA expressions, mutations, and DNA methylation (62)(63)(64). Mutations in the EGFR TK domain have been reported in NSCLC and have been associated with response to gefitinib and erlotinib (65,66). In addition, among the NCI60 cell lines, mutations in the EGFR gene have been identified in the leukemia cell line (RPMI-8226) and melanoma cell line (SK-MEL-28) and were found to be associated with the resistant of these cell lines to 12 tyrosine kinase inhibitors (TKI) including the erlotinib (67,68). It is, therefore, noteworthy that NSC765598 demonstrated high activity on these cells line ( Figure 11, Table 5).
COMPARE analysis indicated that NSC765598 shares a similar antitumor fingerprint with a number of NCI standard agents and a very strong correlation with NCI synthetic compounds (p-value 0.61~0.89). Importantly, the top most correlated synthetic compounds, niclosamide (p-value 0.89) is a known anticancer agent, which have been mechanistically reported to exhibit anticancer activities via suppression of Wnt/b-catenin, mTORC1, EGF, STAT3, NF-kB, and Notch signaling pathways (69,70), while nitazoxanide (p-value 0.62) exhibit it anticancer activities via suppression of MAPK and mTOR pathways (71). The strong correlation of these compounds with the NSC765598 anticancer fingerprint strongly suggests a similar mechanism of action. Hence, it is very likely that anticancer activities demonstrated by NSC765598 could be attributed to suppression of EGF, MAPK, and mTOR signaling pathways, which is in line with the in silico target prediction and molecular docking. Collectively, this study has provided in vitro and in silico evidence for anticancer activities of NSC765598 and inhibition of EGFR, mTOR, NOS2, TGFB1, MAP2K1, and FGFR1 as the most probable mechanism of action. Hence we validated this hypothesis using molecular docking of ligand-protein interactions. Molecular docking of a drug candidate with target proteins is very useful for identifying the strength of an association between a ligand and receptor and ultimately gives a prediction of the activity of the small molecule (18). We conducted a molecular docking study to elucidate likely binding affinities and binding interactions of NSC765598 with selected targets. We found that NSC765598 docked well into the binding cavity of EGFR, mTOR, NOS2, TGFB1, FGFR1, and MAP2K1 with binding affinities ranging −7.2 to −11.0 kcal/mol. Hydrogen bonds together with other non-covalent interactions, such as hydrophobic and ionic interactions and van der Waals forces, play important roles in stabilizing protein-ligand interactions (72). Interestingly, interactions of all targets with NSC765598 predominantly involved conventional hydrogen bonds, fluorine bonds, hydrophobic contacts, van der Waal forces, and a variety of p-interactions including p-p stacking, and p-anion and pcation interactions, which consequently led to high binding affinities and strong ligand-NSC765598 complexes, and likely compromised the expression integrity of the proteins.
Among the targets, EGFR and iNOS demonstrated unique interactions with a higher binding affinity of −11 kcal/mol compared to other targets, thereby suggesting higher promising inhibitory effects of NSC765598 against EGFR and iNOS. The high binding affinities and unique stability of NSC765598 in binding pockets of iNOS and EGFR were chiefly attributed to three conventional H-bonds with TYR483, GLY196, and GLY363 of iNOS and two conventional H-bonds with GLU762 and CYS797 of EGFR; halogen interactions with GLN257, PRO344, and THR184 of iNOS, and ASN842, ARG841, and ASP855 of EGFR; and multiple hydrophobic and Pi interactions (p-stacking, p-alkyl, and p-sulfur). The large numbers of pinteractions, which mostly involve charge transfer, helped NSC765598 intercalate in binding cavities of the receptors. Furthermore, higher Van der Waal forces created on the NSC765598 backbone with the respective amino acids of ILE238, SER236, GLN199, TRP366, MET368, TYR367, ALA345, VAL346, and ASN364 in the binding pocket of iNOS; and ARG858, LYS745, LEU788, and THR854 in the binding pocket of EGFR created strong cohesive environments, thereby stabilizing the complexes formed (73).
The critical role of amino acids in docking has been well demonstrated (74). Alkylation of proteins is dependent on the biochemical properties of target amino acids to create nucleophilic sites (75). Therefore, the specific alkylation of the LEU844 residue of EGFR and alkylation of the three CYS194, LEU203, and PRO344 residues of iNOS by NSC765598 could mediate the modification of substrate binding sites and subsequently induce alkylation-dependent inhibition of target proteins (76). Collectively, the strong interactions observed between NSC765598 and EGFR/iNOS may compromise their activity/expression abilities and consequently affect the survival of cells that depend on these proteins for their vital activities.
These findings, therefore, suggest that the anticancer activities demonstrated by NSC765598 could be attributed more to its dual inhibition of iNOS and EGFR. Specifically, the docking results corroborated our observations of NSCLC cell lines, which were observed to demonstrate higher responses to NSC765598 (GI 50 = 1.12-3.95 µM; TGI = 3.72-16.60 mM). Therefore, stronger NSC765598 interactions with EGFR could be implicated in the higher activity of NSC765598 on NSCLC panels. Interestingly, NSC765598 demonstrated more robust interactions and higher binding affinities for MAP2K, EGFR, iNOS, and FGFR1 than does their respective standard drugs; mirdametinib, gefitinib, N-Iminoethyl-L-lysine dihydrochloride, and erdafitinib. However, it shows, lower binding affinity for mTOR than dactolisib, and its comparable to galunisertib in its affinity for TGFB1. Currently, there are no iNOS inhibitors approved for human use because promising results in animal studies have not translated to humans (77). Therefore, the results of the present study indicate that NSC765598 could serve as a novel iNOS inhibitor worthy of further preclinical and clinical developments.
Hydrophobic interactions play important role in the formation of ligand-receptor complexes. The number of contacts between hydrophobic atoms reflects the extent of the hydrophobic interactions between a ligand and receptor complex. NSC765598 forms several hydrophobic contacts with amino acid residues of the mTOR/EGFR/MAP2K/iNOS/FGFR/ TGFB1 (Tables 3 and 4). These hydrophobic interactions are key players in stabilizing the energetically-favored ligands, in an open conformational environment of the proteins (78). It's noteworthy that NSC765598 forms higher hydrophobic contacts with the receptors than do some of the standard inhibitors used for comparison in this study (Table 4). This higher number of hydrophobic contacts in the active core of the drug -target interface would further increase the biological activity of the drug lead. Collectively, this study suggested that NSC765598 has a potential for multi-target inhibition of EGFR/ iNOS/mTOR/TGFB1/FGFR/MAP2K1 and could serve as a lead compound for developing new therapeutics for cancer treatment. Preclinical toxicity study is an important aspect of drug discovery and developments (79). Our computational study of toxicity indicated that NSC765598 displayed LD 50 at high concentration and, thus may be safely used for acute administration, while the drug-likeness studies indicated that it met acceptable criteria of being used as a drug for therapeutic applications. Further in vitro and in vivo studies in tumor-bearing mice are ongoing to evaluate the full therapeutic efficacy of this novel small molecule.

CONCLUSION
The current study identified mTOR/EGFR/iNOS/MAP2K1/FGFR/ TGFB1 as potential targets of NSC765598 with higher binding preferences for EGFR, iNOS, and mTOR. NSC765598 displayed anti-proliferative activities and selective cytotoxic preferences for NSCLC, leukemia, melanoma, and renal cancer cell lines. In addition, NSC765598 displayed favorable properties as a drug lead compound and thus could be considered a novel small molecule with potential for multi-target inhibition of EGFR/iNOS/ mTOR/TGFB1/FGFR/MAP2K1 and could serve as a lead compound for developing new therapeutics for cancer treatment.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.