In-silico investigations on the anticancer activity of selected 2-aryloxazoline derivatives against breast cancer

Abstract As the in-silico study has become an important tool to search for new drugs in the concurrent era with towering acceptance and accuracy, it has been employed in our research to unearth effective cancer drugs. Breast cancer has accounted for the most serious diseases for both men and women. Although few research outputs have been obtained on breast cancer, these are not an adequate amount to ascertain new drugs. Due to this gap, virtual screening, in-silico study, and computational techniques have been used to provide the ability to design and select anticancer compounds with desirable drug-like properties of breast cancer protein, which is commonly known as fatty acid synthase. A total of nine derivatives of 2-aryloxazoline compounds were chosen, and In-silico was studied to evaluate as a potential anticancer agent with the comparison of seven Food and Drug Administration(FDA) approved breast cancer drugs. These compounds were subjected to computational studies for quantum calculations, ADME and Lipinski analysis, as well as molecular docking and MD simulations against a variety of therapeutic targets involved in cell proliferation of fatty acid synthase (PDB ID:3TJM, 3ERT, 4OAR, 2J6M). An in-silico docking study reveals that ligands Hit-4, Hit-6, and Hit-8 had the highest docking scores at −10.3 kcal/mol, −10.3 kcal/mol, and −10.2 kcal/mol towards the protein of fatty acid synthase. The ligands had docking scores better than the standard anti-breast cancer drug gefitinib (−5.3 kcal/mole). Our findings demonstrate how crucial it is for pharmaceutical researchers to develop novel drugs for the treatment of breast cancer. Communicated by Ramaswamy H. Sarma


Introduction
According to WHO statistics [2012], breast cancer has become one of the second major causes of death in Western countries, behind only AIDS [2010]. It now accounts for 23% of all cancer deaths in Asia and is the second leading cause of death in Western countries (Donepudi et al., 2014). More than 60% of breast cancer cases in Asian countries are estrogen receptor alpha-positive (ER) cancers. ER plays a crucial role in breast cancer development in normal mammary gland development. However, in ER breast cancer cell lines in vitro, ER specifically mediates the proliferation of estrogen-induced cells in an autocrine mode of action (Tan et al., 2009). The progesterone hormone receptor and mammaglobin in a stepwise approach are the breast cancer diagnostic tools with the highest positive predictive value (88%) with estrogen (de las Mercedes Noriega et al., 2012). During puberty, estrogen concentrations rise, stimulating the synthesis of estrogen and progesterone receptors in the mammary glands. These sex hormones can contribute to the development of breast cancer (Kuhl & Schneider, 2013). On the other hand, an increase in breast mass is connected with an increase in the number of fat cells that produce cholesterol estradiol, which enhances the risk of breast cancer (Akram et al., 2017).
Breast cancer chemotherapy is characterized by the targeting of receptor functions such as ER, PR (progesterone receptor), EGFR (epidermal growth factor receptor), and others. It's critical to comprehend the involvement of ER in breast cancer growth and progression. Anti-estrogen therapy, which was the first targeted therapy for breast cancer in humans, is a good way to treat ER-positive breast cancer (Wang & Yin, 2015). Overexpression of PR in breast cancer is common and is directly linked to ER over-expression. PR and ER overexpression improve the prognosis of PR-positive breast cancer and increase the likelihood of response to hormonal therapy (Kiani et al., 2006).
Breast cancer treatment is hindered by two major obstacles: metastasis and drug resistance, both of which are linked to fatty acid metabolism. When compared to normal cells, tumor cells have a wide range of significant differences in their metabolic activity. Changes in fatty acid metabolism are an important part of the epithelial-to-mesenchymal transition, invasion, and spread of tumor cells (Cuy� as et al., 2014). FASN (fatty acid synthase) is a key enzyme in endogenous fatty acid synthesis. Most normal cells don't make much FASN, but this enzyme is often found in breast cancer cells (Smith, 1994). Inhibiting FASN may reduce tumor cell proliferation and survival while also inducing apoptosis. FASN expression has also been linked to tumor grade, cancer aggressiveness, and a poor prognosis (Jensen et al., 1995;Kuhajda et al., 1989;Menendez & Lupu, 2007;Milgraum et al., 1997). So, blocking FASN could be a good way to treat breast cancer that has spread to other parts of the body.
Oxazoline and its derivatives have been found to have pharmacological activities (Bansal & Halve, 2014). Many oxazoline derivatives have been reported to have anticancer activity (Gros et al., 2012;Li et al., 2002). As a consequence, newly synthesized oxazoline derivatives can be used as desired ligands in virtual screening methods. 2-Aryloxazoline scaffolds have been the basis for many biologically active molecules that are cytotoxic, antibacterial, antitumor, antidepressant, and anti-Alzheimer's (Celanire et al., 2011;Garg et al., 2014;Kumar et al., 2019;Onishi et al., 1996;Tsuda et al., 2005;Tsukamoto et al., 1997;Vizi, 1986). These moieties are also used as ligands in a variety of organic transformations. Many polyoxazolines are also useful biomaterials for drug and gene delivery, as well as stimuli-responsive systems. Furthermore, oxazolines can be dehydrogenated to form oxazoles, which are another important class of bioactive molecules. The antioxidant properties of aryl-substituted 2oxazolines have recently been investigated (Djurendic et al., 2011;Padmaja et al., 2014;Padmavathi et al., 2010). 2-Aryl oxazoline compounds also show antifungal activity with low cytotoxicity. Our present in-silico studies of 2-aryloxazoline compounds, which were synthesized by Luis M. Z. Argomedo et al. (Scheme 1), are shown below (Argomedo et al., 2020).
Considering the role of specific target receptors such as Fatty Acid Synthase (FASN), ER (Estrogen receptor), PR (Progesterone receptor), and EGFR (Epidermal growth factor receptor) in the initiation and progression of breast cancer, we have proposed nine 2-aryloxazoline derivatives (Hits 4, 6, and 8) in Figure 1 with anti-cancer potential.

Results and discussion
The goal of the study was to elucidate alternative inhibitory compounds against Fatty Acid Synthase (FASN), Estrogen Receptor Alpha (ERA), Progesterone Receptor (EGFR), and Fatty Acid Synthase (FASN). Scheme 1 was used to do a virtual screening of 2-aryloxazoline compounds in Table S1 to find ones that work well against breast cancer.
Reasons for selecting specific proteins

Molecular docking
The binding affinities of various 2-aryloxazoline derivatives with some common protein targets for breast cancer were studied in this work. The value was calculated using the binding energies in Table 1. A total of nine 2-aryloxazoline compounds were confirmed as ligands based on their binding affinities with the receptor targets Fatty Acid Synthase, ER, PR, and EGFR. Nine compounds were found to exhibit a strong dock score with the Fatty acid synthase (FASN) and a modest dock score with additional target proteins. Hits 4, 6, and 8 have shown the highest dock score À 10.3 kcal/mol and À 10.2 kcal/mol against the FASN (PDB id: 3TJM). Hit 4 has shown zero, four, two, and one hydrogen bonds with the active site amino acids; PDB ids: 3TJM, 3ERT, 4OAR, and 2J6M, respectively, with dock scores of À 10.3 kcal/mol, À 8.2 kcal/mol, À 8.7 kcal/mol, and À 8.2 kcal/mol. Hit 4 was found to form stable p-sigma, p-p stacked, p-alkyl bonds with the active site's residues LEU2427, TYR2351, and PHE2370. Hit 6 has zero, two, and three hydrogen bonds with the respective receptors, yielding dock scores of À 10.3 kcal/mol, À 8.1 kcal/mol, À 8.7 kcal/mol, and À 7.8 kcal/ mol. Hit 6 established stable p-sigma, p-p stacked, p-alkyl bonds with the LEU2427, TYR2351, and PHE2370 residues in Figure 2. Hit 8 has formed one, five, two, and one hydrogen bonds with the dock scores of À 10.2 kcal/mol, À 8.7 kcal/mol, À 8.7 kcal/mol, and À 8.0 kcal/mol respectively. The H-bond was formed between halogen and the ALA2367 where halogen acts as H-acceptor. This compound also formed p-p

PDB ID Functions 3TJM
Acetyl-CoA metabolic process, fatty acid biosynthetic process, fatty acid metabolic process, fatty-acyl-CoA biosynthetic process, glandular epithelial cell development, positive regulation of cellular metabolic process, regulation of lipid metabolic process, dehydratase activity, hydrolase activity, RNA binding (Jayakumar et al., 1994). Fatty acid metabolism is linked to the occurrence and progression of tumors. Thus, 3TJM is a potential target for our small molecule. Continued.

PDB ID Functions 3ERT
Negative regulation of DNA-binding transcription factor activity, negative regulation of gene expression, androgen metabolic process, cellular response to estradiol stimulus, epithelial cell proliferation involved in mammary gland duct elongation, intracellular estrogen receptor signaling pathway, positive regulation of fibroblast proliferation, regulation of transcription, DNA-templated, regulation of transcription by RNA polymerase II, response to estrogen , estrogen receptor activity (Lagani� ere et al., 2005;Matsuda et al., 2002;van de Stolpe et al., 2004). Since hormonal therapy is popular in the treatment of breast cancer, we have chosen this protein (3ERT) as our targeted protein.

4OAR
Signaling between cells, intracellular steroid hormone receptor signaling pathway, negative regulation of gene expression, positive regulation of transcription by RNA polymerase II, progesterone receptor signaling pathway, transcription initiation from RNA polymerase II promoter, ATPase binding (Yin et al., 2007). Progesterone has been linked to breast cancer because it is a growth hormone in a normal breast and because it is a risk factor for breast cancer over a lifetime (Daniel et al., 2011). So here, this specific protein (4OAR) has been chosen for research.

2J6M
Positive regulation of cyclin-dependent protein serine/ threonine kinase activity, epidermal growth factoractivated receptor activity, protein kinase binding, positive regulation of epithelial cell proliferation (Andl et al., 2003). The epidermal growth factor receptor (EGFR) plays an important role in the development and progression of many cancers including breast cancer (Normanno et al., 2006). For this, 2J6M was selected as a research target protein.
stacked, and p-alkyl bonds with TYR2351, PHE2370, and LEU2427 respectively. In silico studies of protein-ligand interactions revealed that 2-aryloxazoline derivatives play an important role in the formation of H-bond and interactions. The docking interactions for all the Hit 1-9 were also shown in Table S8. The binding affinity in Table 2 of seven FDA-approved drugs in Table S2. Drug one has the highest dock score among the seven drugs, with a dock score of À 11.2 kcal/mol against Fatty Acid Synthase (FASN) PDB id: 3TJM. Drugs two, four, and six had the best dock scores against Estrogen receptor alpha (PDB id: 3ERT): À 9.8 kcal/mol, À 7.3 kcal/mol, and À 8.0 kcal/mol, respectively. The dock score of drugs three with the Epidermal growth factor receptor (PDB id: 2J6M) is À 9.7 kcal/mol. Drug five has the best dock score of À 9.8 kcal/mol against the Progesterone receptor (PDB id: 4OAR). Drug seven has the best docking score of À 8.6 kcal/ mol when interacting with the Fatty Acid Synthase (PDB id: 3TJM) (Figure 3).

Prime MM-GBSA
where DG(solv) is the difference in GBSA solvation energy of the protein-ligand complex, DE(MM) is a difference in the minimized energies between protein-ligand complex, and DG(SA) is a difference in surface area energies of the complex. Calculations based on MM/PB(GB)SA have been useful in determining the most reliable binding mode of ligands. To accomplish this, various ligand orientations generated by docking software are used as starting ligand-receptor coordinates for MD simulations, which are then processed with the MM/PB(GB)SA evaluation. The most dependable ligand orientation is then identified based on the best binding free energy obtained (Kumari et al., 2014;Tuccinardi, 2021) The prime MM-GBSA analysis revealed the binding energy DG of FDA-1 to 3TJM as À 42.88 kcal/mol compared with best-docked Hits 4 and 6 with respectively À 51.15 and À 59.52 kcal/mol.

Toxicity risk assessment screening
The 2-aryoxazoline compounds were tested for toxic properties such as irritant, mutagenic, tumorigenic, and reproductive effects using Molinsperation and the Osiris server (Sarkar et al., 2021). The server includes a list of approximately 5300 distinct substructure fragments derived from 15,000 commercially available fragments, each with a reported drug score and drug likeness. To assess the overall potential of a compound to qualify as a drug, the drug score related with drug likeness, cLogP, molecular weight, and toxicity risks may be added together to form a total value. Toxicity testing on all 2-aryoxazoline compounds except Hit 9 revealed no risk of mutagenic, tumorigenic, irritant, or reproductive toxicity without one compound in Table 4. And chemical reactivity is found in Tables S3 and S5.

Lipinski's rule violations
The ADME was used to determine the bioavailability of (Hit 1-9) 2-aryloxazoline compounds (Adsorption, Distribution, Metabolism, and Excretion) using the Molinspiration server (https://molinspiration.com/) and Swiss ADME. To investigate the drug-like properties of Hit 1-9, the lipophilicity, expressed as the octanol/water partition coefficient and referred to here as log P(o/w), as well as other theoretical calculations such as molecular size, TPSA, the number of hydrogen bond acceptors and donors, and the number of rotatable bonds, were calculated. Lipinski's fifth criterion is used to assess a chemical compound's drug similarity if it has characteristics that would make it a probable or possible drug in humans (Lipinski et al., 1997). Certain molecular parameters, such as log P, a polar surface, the number of donors of the hydrogen bonds, the number of receivers of hydrogen bonding and molecular weight, are used in the prediction of the oral activity of a drug substance. It has been established that the most complexes with good membrane permeability have log P � 5, number of hydrogen bond acceptors � 10, and number of hydrogen bond donors � 5. In most cases, an orally active medication fits all of the criteria. The ligands in this study were found to be consistent and oral bioavailable with the criteria in Table 5.

Bioactivity score of the ligand
The term pharmacological activity  refers to the beneficial effects of drugs on living organisms. The drug is meant to bind to a biological target. The most common biological targets are proteins such as enzymes, ion channels, and receptors. Bioactivity scores have been determined with various parameters in selected compounds, including GPCR binding and ligand nuclear receptors, ion channel modulation, and inhibition of kinases, protease inhibition and inhibition of enzyme activity. Table 6 shows the bioactivity score. It is known that if the bioactivity score is greater than 0.0, the complex is active; if it is between À 5.0 and 0.0, the complex is moderately active; and if it is less than À 5.0, the complex is inactive. As shown in Table 6, the ligands' bioactivity scores ranged from À 5 to 0, which shows that they have the qualities needed for the complexes to work as possible drugs (Dar et al., 2016).

Lipophilicity indices
Control of lipophilic efficiency indexes, such as ligand lipophilic efficiency LipE, and ligand-efficiency dependent lipophilicity (LELP), have the potential to make a major contribution to overall drug quality at various stages of the drug discovery process. If the lipophilicity is too high, there is a greater chance of binding to targets other than those desired, and thus there is a greater risk of toxicity (Leeson & Springthorpe, 2007). To make it easier to optimize the affinity in terms of lipophilia, the efficacy of ligand lipophilicity efficiency has been demonstrated. Leeson and Springthorpe (2007) established LipE (LLE) as a parameter that seeks to improve potency while maintaining low lipophilicity, hence increasing the specificity of the interaction with the receptor during the contact (Table 7 and Table S4).
Here, pIC 50  (Wager et al., 2010). Hit-4 and 7 have value and have been deemed to be the best of their kind. When attempting to achieve ideal ADMET characteristics, the size of the molecules and their lipophilicity are important considerations. The ligand efficiency-dependent lipophilicity index (LELP) proposed by Keseru and Makara (2009) combines molecular size and lipophilia into a single measure of efficacy. The ideal LELP score is À 10 < LELP < 10. LELP ¼ Log P/LE Hit-4 and 7 have LELP scores of 7.888 and 7.915, respectively, which fall within the suggested score range of À 10 < LELP < 10. On the other hand, Hit-1, 2, 6, and 8 have LELP less than 16.5, indicating that they are in the Lipinski zone (ROF-score ¼ 4). However, Hit-3, 5, and 9 have LELP scores of 17. 16, 20.49, and 17.47, respectively, which is consistent with their low weak ROF score < 4. The supporting Table S6 shows the side effects of approved drugs are the primary reason for looking for new drugs with fewer side effects. Thus, as a potential alternative to these drugs, we have proposed new compounds.

Molecular docking analysis
According to molecular docking analysis of 2-aryloxazoline compounds and FDA approved drugs, both have the highest dock score against Fatty Acid Synthase. One approved drug had a dock score greater than À 10 kcal/mol, while three 2aryloxazoline derivatives had a dock score greater than À 10 kcal/mol. Overall, approved drugs and our chosen hit compounds had comparable dock scores with the four target proteins.

Cheminformatics analysis
We have performed cheminformatics analysis on the compounds to compare the approved drugs with our proposed drugs. Compounds were analyzed for our pre-defined set of structural and physicochemical parameters (Stratton et al., 2015). According to Table 5, approved drugs one, three, and seven violate Lipinski's rule by one, three, and one, respectively, whereas our proposed drugs in Table 4 do not have any violations of the rule. Drug One violates the rule by having a molecular weight of more than 500. Drug three, on the other hand, is in violation of the rule since its molecular weight is greater than 500, its TPSA is greater than 140, and its HBA is greater than 10. Drug seven has a Log p violation that is greater than 8. Subsequent studies found a link between increased oral bioavailability with rotatable bond count (Rot B � 10) and topological polar surface area (TPSA � 140).
Another significant aspect of molecular complexity was defined by Lovering as the fraction sp3(Fsp3), where Fsp3 ¼ total number of sp3 carbons � total carbon count. Importantly, Fsp3 has been linked to faster progression from lead discovery to clinical trials and drug approval (Lovering et al., 2009). The Fsp3 values of approved drugs in Table 5 do not differ significantly from those of proposed drugs in Table 4. This is especially significant because increased molecular complexity, as measured by Fsp3, has been linked  to molecules' ability to investigate larger regions of chemical space. As a result, the molecular complexity of approved and proposed drugs is comparable.

Principal component analysis comparison of compound classes
Principal component analysis (PCA) is an authoritative and trendy multivariate investigation process with quantitative variables in biostatistics, chemio-informatics, bioinformatics, and many other fields, as well as an important and useful arithmetic tool in drug design and discovery. In an emblematic quantitative structure-activity relationship (QSAR) study, PCA analyzes an original data matrix in which molecules are depicted by several intercorrelated quantitative dependent variables (ADME and physiochemical descriptors). In Figure 8, the validation percentage of tested compounds is represented with two changeable variables, such as PC1 and PC2. With the Y axis, 23.3% validity is obtained, and 45.9% is illustrated on the X axis. In addition, Hits 3 and 5 convey the maximum validation in terms of drug-likeness and the Lipinski Rule.

Protein structure preparation
In this research, the structure of Fatty Acid synthase receptor (FASN) (PDB id: 3TJM), Estrogen receptor alpha (PDB id: 3ERT), Progesterone receptor (PGR) (PDB id: 4OAR), and Epidermal growth factor receptor (EGFR) (PDB id:2J6M) were retrieved from the protein data bank, RCSB. The PDB structure contains a number of gaps in specific connectivity information, as well as bond orders and formal changes. Thus, pre-process of protein is required. Pre-processing is important for subsequent structure preparation actions such as generating heteroamorous states, assigning H-bonds, and minimizing the level of minimization

Ligand selection and preparation
We have selected 2-aryl oxazoline compounds based on their minimum inhibitory concentration and antifungal activities. These compounds have been used as ligands for docking analysis. 2-aryloxazoline compounds that are already synthesized were taken for ligand preparation. Total nine compounds were taken for ligand preparation. All of the molecules were drawn in sdf format using the 2D and 3D options of Chem. Draw Ultra 16.0. The Gaussian 16 W and Gauss View performed basis set 3-21 G (d, p) were used for energy minimization, conformational analysis, and ligand preparation. These molecules were then imported into the docking library of the Auto dock (version 4) docking software and docked.

Molecular docking
The docking of 2-aryloxazoline compounds was done with their targeted receptors by using Autodock 4.00 Vina. To identify the correct ligand for the target both programs are used. The best three conformation of the substrate were taken to the next step for molecular dynamic study based on their docking score. For ligand docking in the protein and their interaction visualization, open-source software AutoDock Vina, PyRx, Discovery Studio, and PyMol programmes were used.

Molecular dynamics simulation
For molecular docking simulation, three selected L1-Co, L2-Co and L1-Cu metal complexes were performed by the YASARA dynamics suite using the AMBER14 force field. The process was fixed by assembling NaCl salt at 0.9% concentration, pH 7.4 at 298 K (25 � C) temperature, and water molecules. Lastly, a 50 nanosecond(ns) MD simulation was executed, with snapshots taken every 100 ps along the MD trajectory for additional study. The RMSD and RMSF values were examined for each simulation run. MD snapshots were raised to evaluate their interaction. For all the prediction and simulation studies, we used the HP Envyx360 (intel core i7) laptop.

Toxicity, bioactivity and lipophilicity indices prediction server
The Molinspiration server (https://molinspiration.com), Osiris, and Swiss ADME software were used to calculate the toxicity and bioactivity of the compounds under consideration. With the help of the ChemDes software, numerous components were computed in order to determine pIC50.

Conclusions
The current study involves an in-silico study for the efficacy and convenience of nine derivatives of 2-aryloxazoline and seven FDA-approved breast cancer drugs. This work focused on the anticancer activity of 2-aryloxazoline derivatives against the target proteins that are responsible for causing breast cancer and mostly focused on inhibitory activity against Fatty Acid Synthase (FASN). In order to identify compounds with high potency, they were evaluated in comparison to the approved drugs. Using Lipinski's rule, lipophilicity indices, and bioactivity score techniques, it was found that Hits 4, 6, and 8 have much better oral bioavailability and in silico ADME properties. In addition, these compounds have safety characteristics with low toxicity and non-carcinogenic properties. Next, the molecular docking against fatty acid synthase, estrogen receptor alpha protein, progesterone receptor, and epidermal growth factor receptor where the highest docking values are received for the protein of fatty acid synthase from À 9.0 kcal/mol to À 10.3 kcal/mol. Hit 6 and Hit 8 show the highest docking values of 10.3 and À 10.1 kcal/mol, respectively, for fatty acid synthase, and Hit 5 illustrates the highest docking score against the other three proteins. The approved drugs also meet most of these requirements, which means that Hit-4, 6, and 8 are more likely to experience positive results in the clinic. It can be said that the docking value of best Hits is almost greater than that of approved drugs. Researchers in the future will be able to use these results to make and test new anti-cancer drugs that target Fatty Acid Synthase (FASN).

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
The author(s) reported there is no funding associated with the work featured in this article.

Data and software availability
AutoDock Vina, PyRx, Discover Studio, PyMOL software and web server used in this analysis is free for academic use. And licensed software's are Gaussian16, Yet Another Scientific Artificial Reality Application (YASARA) used for Molecular Dynamics simulation.