An In silico Approach to Identify High Affinity Small Molecule Targeting m-TOR Inhibitors for the Clinical Treatment of Breast Cancer

Breast cancer is the most frequent malignancy among women. It is a heterogeneous disease with different subtypes defined by its hormone receptor. A hormone receptor is mainly concerned with the progression of the PI3K/AKT/mTOR pathway which is often dysregulated in breast cancer. This is a major signaling pathway that controls the activities such as cell growth, cell division, and cell proliferation. The present study aims to suppress mTOR protein by its various inhibitors and to select one with the highest binding affinity to the receptor protein. Out of 40 inhibitors of mTOR against breast cancer, SF1126 was identified to have the best docking score of -8.705, using Schrodinger Suite which was further subjected for high throughput screening to obtain best similar compound using Lipinski’s filters. The compound obtained after virtual screening, ID: ZINC85569445 is seen to have the highest affinity with the target protein mTOR. The same result based on the binding free energy analysis using MM-GBSA showed that the compound ZINC85569445 to have the the highest binding free energy. The next study of interaction between the ligand and receptor protein with the pharmacophore mapping showed the best conjugates, and the ZINC85569445 can be further studied for future benefits of treatment of breast cancer.

Introduction novel treatment to prevent breast cancer. The top-most encouraging efforts require understanding the gene, gene-environment, or gene-gene interaction.
The leading cause of cancer is the genetic mutation which occurs as a result of genetic instability and environmental factors. In most of the breast cancer, genetic alteration arises during an individual's lifetime and are present only in specific cells of the breast. These types of changes are called somatic mutation and are not inherited. A different kind of genetic changes, which is classified as germline mutation, are typically inherited from parent to their offspring.
Mutation in the number of genes is involved in causing breast cancer. Regarding high-risk family history, the most important genes are BRCA1 and BRCA2 (Ford et al., 1998). Variation in these two genes can increase the chance of developing the tumors by 80%. Therefore, early recognition of carrier among affected women is pivotal to suggest the specialist to determine the best suitable treatment plan (D'Argenio et al., 2015). mTOR pathway mTOR kinase is an intracellular signaling pathway which controls the various actions of our body such as cell division, cell survival, cellular morphology, protein synthesis and integration of metabolism. mTOR kinase signaling is activated in multiple cancer, guided by variation in the gene-coding receptor tyrosine kinase, Ras, PI3K, and PTEN that is involved in numerous cellular processes (Bhagwat et al., 2011). This protein is a component of two multi-subunit complexes known as mTOR1 and mTOR2. The mTOR1 complex is sensitive to anticancer drug Rapamycin, an allosteric inhibitor of mTORC1 results in complete disruption of the complex. The mTOR2 complex is considered as rapamycin-insensitive, regulates actin cytoskeleton by phosphorylating the survival kinase AKT at SER 473. mTOR is mainly activated by some intracellular and extracellular signals such as growth factors, nutrient status, energy metabolism and Oxygen level (O'Reganet al., 2011;Aarthy et al., 2017).

Dysregulation of the mTOR pathway
The aberrant activation of mTOR pathway identified in most of the cancer including breast cancer and its hyperactivation commonly associated with cell growth, cellular proliferation and neogenesis (García-Echeverría et al., 2010). For instance, in the variety of cancer phosphatase and tensin homologue deleted in chromosome 10(PTEN) is mutationally inactivated, leads to an increase in mTOR activity. This hyperactivated mTOR gene results in the production of mRNAs that encode growth factors, cell death inhibitor, angiogenesis factors, cell growth inducer which overall promote carcinogenesis. Therefore, mTOR must be specifically targeted as an anticancer therapy for the treatment of cancer (Dowling et al., 2007). Overall features of the mTOR signaling pathway have provided a higher level of interest in targeting mTOR as a potential therapeutic agent for effective treatment (Xie et al., 2016).

System Configuration
The present study was performed using a Schrödinger's Drug Discovery Suite 2015 on the platform of Cent OS Linux 6.5 version. Molecular Dynamics simulations were carried out using an academic version of Desmond 2015.

Selection of Protein Target and known inhibitors
The three-dimensional structure of the target protein mTOR was retrieved from Protein Data Bank -PDB (www. rcsb.org) with PDB ID: 5H64. Established inhibitors for mammalian target of rapamycin (mTOR) were explored using NCBI's PubChem compound database (Chandrakar, B. et al., 2013;Rao DM. et al., 2010). The total of 40 compounds were selected to consider the best-anchored compound. The selected inhibitors along with their PubChem ID are shown in Table 1.

Lead Compounds and Database preparation
All the 40 compounds retrieved from NCBI's PubChem Database and a large chemical library of drug-like compounds -Zinc databases were prepared using the Ligand preparation module (LigPrep, 2017) (Dunna et al., 2015;Dunna et al., 2015; of molecular modeling package with suitable parameters like optimization, ring conformation, 2D to 3D conversion, determination of protomers, tautomers, and ionization states at pH 7.0, along with partial atomic charges using OPLS_2005 force field (Bandaruet al., 2014;Jorgensen et al., 1996;LigPrep, 2015;Sinha et al., 2015).

Active Site Prediction and Receptor Grid Generation
Using SiteMap module in Schrodinger Suite, Ligand -Binding site of the receptor was predicted. Based during clinical studies due to poor ADME/T; it leads to high cost of loss to pharmaceutical companies. Thus, QikProp, Schrodinger was taken forward for profiling the ADME properties of selected hits from screening. It evaluates the drug likeness and pharmaceutical properties such as molecular weight, aqueous solubility (Q P log S ), octanol/water (log P ), brain/blood partition coefficient (Q P log BB ), CNS, hydrogen bond donors and acceptors along with Lipinski rule of five and Jorgensen rule of three. Following to these, the Human intestinal absorption (HIA), Blood brain barrier (BBB), AMES toxicity and LD50 were also generated to check compounds toxicity effect using online web server tool admetSAR (QikProp 2015;Cheng et al., 2012).

MM-GBSA (Molecular Mechanics, the Generalized Born model and Solvent Accessibility) -free binding energy calculation
MM-GBSA, an efficient computational approach in Prime module of Schrodinger suit 2015 (Schrodinger, Inc., LLC, New York, USA), is worthwhile to calculate the relative binding free energy and to improve docking score after docking analysis from bio-molecular system. The binding free energy determines by ΔG bind , represented by following equation (Prime, Schrodinger, 2015;Lyne et al., 2006): The Prime module applies a surface generalized Born model which makes use of a Gaussian surface for enhanced demonstration of a solvent accessible surface area (Suryanarayanan et al., 2005;Singh et al., 2016;Prime, Schrodinger, 2015;Lyne et al., 2006).

BOILED-Egg plot
A Brain or intestinal EstimateD permeation method (BOILED-Egg) is a predictive model which predicts the bioavailability of the drugs over gastrointestinal absorption and brain penetration. There are five parameters which defines the Cartesian coordinates of both ellipses includes MW, TPSA, MLOGP, GI and BBB. Classification of Egg-plot include the yolk (physiochemical space for BBB permeation), the white (physiochemical space for HIA absorption), and the outside grey region stands for the molecule which predicted low absorption and limited brain penetration properties. The molecule which are placed in grey region are counted as remarked ( on top ranked site score, five potential active sites were analyzed. Site2 was taken for docking analysis based on highest site score along with presence of hydrophobic and charged amino acids (Sinha et al., 2014;Reddy et al., 2014;Panwar et al., 2017).Using Receptor Grid Generation module in Schrodinger, the grid generation was performed on prepared protein with the help of site score from SiteMap (Bandaru et al., 2014;Bandaru et al., 2013). The atoms of protein were fixed within the default parameters of the radii of Vander Waal's scaling factor of 1 Å with partial charge cut-off of 0.25Å using OPLS_2005 force field (Vuree et al., 2013).
Based on the glide docking score the best-docked complex was selected for further studies of virtual screening (Suryanarayanan et al., 2005;Singh et al., 2016).

High throughput Virtual Screening
Virtual screening is one of the highly advanced computing processes to identify the novel drug candidate from large chemical libraries against biological target. Herein, the trio of HTVS, SP, and XP from Glide, Schrodinger was utilized to filter the best configuration of ligand with highest docking score in every step of virtual screening trio (Aarthyet al., 2017;Suryanarayanan et al., 2005;Singh et al., 2016). Finally, the best ranking compounds were selected using criteria of docking score for further studies of ADMEprediction.

ADME and Toxicityprofile
As per olden research, many of drugs have been failed To proceed for further analysis of Egg-plot, compound ID of the top three best drug from each of established docked and virtual screened docked was retrieved. Observable result were analyzed based on the different parameters used for Egg-plot.

Results
Herein this study, a highly significant virtual screening process was applied to find out an effective leading compound using Schrodinger Software.

Docking and virtual screening results
Molecular docking of top 40 known inhibitors was identified with great binding interaction within the same active site. Later, SF1126, the best one protein-ligand complexes were analyzed as perfect binding confirmation with better docking score, XP Score, glide energy, and Glide emodel, tabulated (  (Table 3). The resulted compound ID:ZINC85569445shows the best affinity with the target protein with docking score of -10.607 kcal/mol. The binding mode analyses of the ZINC85569445 were described in detailed.

Binding mode of Compound ZINC85569445 with the receptor
The ligand ZINC85569445was identified with highest docking score -10.607 kcal/mol, glide energy -61.060 kcal/mol, glide Emodel -82.947 kJ/mol. Hydrogen bond interactions were identified with the amino acids Glu662, Lys690, Glu701, Asn725; in which, amine group of compound interacted with oxygen of Glu662 with a distance of 2.01 Å, three carboxyl group of compound interacted respectively with amino group of Lys690 with a distance of 2.43 Å, oxygen of Glu701 with a distance 1.96 Å, and oxygen of Asn725 with a distance 2.07 Å. Amino acids residues Val671, Tyr674, Phe678, Val681, Leu694, Pro697, Ala698, Tyr723, Ala732, Leu735, Leu742 were observed as hydrophobic residues. The 2D profile interaction diagram was represented in Figure 1.

Pharmacophore Studies
Pharmacophore mapping helps to understand the interaction between ligand and receptor molecule. In the active site of target protein, compound ID:ZINC85569445 shows considerable interaction. It shows electrostatic interaction with Ser722 and Lys690, while the H-bond was observed with Glu662, Lys690, Glu701, and Asn725. Figure 4, depicts the aromatic interaction where Phe678, Thr731 actively participated.

Hydrogen Bond interaction between compound ID and target protein
ZINC85569445 and the target protein mTOR is shown (Figure 2). Green dotted lines represent the Hydrogen interaction between atoms. This interaction involving atoms of the residues Asn725, Glu701, Val67, Ser722, Lys690 of mTOR.   Electrostatic Interaction between the compounds ID ZINC85569445 in the active site of mTOR shown in (Figure 3). The red surface of the protein is electrically negative surface;while, the blue surface is electrically positive. The compound is deeply embedded in the cavity of positive and negative amino acids of the target protein mTOR.

Aromatic interactions between the most effective compoundID
ZINC85569445 and mTOR protein shown in ( Figure  4). Most favorable region of the aromatic interaction is the blue region (Edge), where Phe678 of mTOR is actively participating in bond formation. While the least favorable region (Face) have less interaction.

Interactions of the compounds ID
ZINC85569445 in the active site of mTOR shown in [Fig 5]. The residues (Asn725, Glu701, Val67, Ser722, and Lys690) of blue dotted line participating in Hydrogen bond interaction while the residues in green circles forming Van-der Waals. Pink circled residues participating in electrostatic interaction.

ADME profile
An ADME property of the top 10 hits was calculated by evaluating their physicochemical properties using QikProp, Schrodinger. All the predicted ADME properties are accepted within the standardized range defined for human use such as Molecular weight (MW 130-500), H-Bond donor (< 5.0), H-Bond acceptors (< 10.0), the octanol/water partition coefficient log (-2 to 6.2), the aqueous solubility log (mol/L) (-6.5 to 0.5), CNS activity -2 (inactive) to +2 (active Lipinski's rule of five and Jorgensen rule of three), shown in Table 4.

Comparative ADMET profile of the test ligands and the control
Comparative studies of the parameters such as Human intestinal absorption(HIA), Blood-brain barrier (BBB), AMES toxicity, LD50 doses of best established docked compound SF1126 (control) and top three virtual screened ligands (CID: ZINC85569445, CID: ZINC14640443, and CID: ZINC85489178) was done: using R-programming language which is depicted in Figure 6. ADMET parameters such as HIA, BBB, AMES toxicity, and LD50   compound. CID: ZINC85489178 had almost equivalent LD50 dose to SF1126. Regarding toxicity, the probability of being toxic was lower for SF1126 and ZINC85569445.
Ames test employed to test whether a compound is toxic or not. The result of the Ames test revealed that all ligands and control molecule were non toxic.

MM-GBSA
The binding free energy was calculated using a post-scoring method -MMGBSA for the evaluation of molecular docking process. Obtained MMGBSA (ΔG bind) ranges from -58.149 to -89.989, are presented in Table 4. Results were correlated along with docking score to design a relevant drug like potent inhibitors. Higher binding free energy of the lead defines the greater affinity to bind with the receptor.

BOILED-Egg plot analysis
To understand the poor pharmacokinetics behaviors and bioavailability of the drug which is crucial for drug discovery process, BOILED-Egg plot is proposed as an accurate model to predict lipophilicity and polarity of the small molecule. Prediction for both BBB and intestinal permeation is translated into a molecular design on    (Figure 7).

Discussion
Breast Cancer is considered as the prime cause of mortality among women worldwide. Metastatic breast cancer is a diverse disease with different subtypes and irremediable with current treatment regimens (Brouckaert et al., 2017). Based on the data composed by American cancer society, nearly 231,840 recent cases supposed to be diagnosed in 2105 (Ward et al., 2015). One important cause of cancer is the sequential mutation in the number of genes due to the fact of genetic instability and environmental factors (Al-Hajj et al., 2003). mTOR (mammalian target of rapamycin) is a serine/threonine kinase, which controls the various acts of our body such as cell growth, survival, metabolism, is upregulated in various cancer. mTOR is a downstream member of the PI3K/AKT signaling pathway, and its aberrant activation is observed in various human malignant diseases (Vicier et al., 2014). Therefore, mTOR is considered as a desirable target for treatment strategies.
Accordingly, substantial importance is given on PI3K/ AKT/mTOR signaling pathway which has led to the progress of different kinds of inhibitors. In the present study, we found a pre-existing drug against mTOR and technically found SF1126 with the best binding score -8.705kcal/mol against mTOR. Further exploration established from the structure similarity search of SF1126 against ZINC database found various compound with similar properties against mTOR. High-throughput virtual screening performed in this study for structure similarity search which revealed ZINC85569445 with best binding score -10.607kcal/mol. In this study the tested compound CID: ZINC14640443 displayed with the lowest binding energy of -10.607kcal/mol. The binding energy of the control molecule SF1126 was higher than CID: ZINC14640443 as found in our study, thus CID: ZINC14640443 displayed the much better binding score. ADME profile of both the compounds SF1126(control) and test ligand ZINC85569445 was evaluated using QikProp Schrodinger and all the parameters accepted within the standardized range defined for human use. CID: ZINC14640443 predicted blood-brain barrier 0.51 compare to SF1126 of 0.50. Carcinogenic profile of both the test ligand and control molecule revealed to be non-carcinogenic.
The effectiveness of the known inhibitors against mTOR has been tested with significant antiproliferative activity while some of them are being under phase II/III clinical trials (Chresta et al., 2010). In spite of having high ability in inhibiting the activity of mTOR, inhibitors are quite inadequate in fighting against cancer. The likelihood, there is the number of reasons such as feedback loop can activate the upstream signaling pathway and promote cell survival and proliferation. The mTOR signaling pathway is essential for healthy cell growth and viability and its inhibition can be destructive to other cells and tissues (Xie et al., 2016). Despite some loopholes, further attention should be given to the understanding of the mTOR signaling pathway and its downstream processing which play a vital role in tumor progression.
In conclusion, this research study focuses on finding a satisfactory lead compound which targets to inhibit the breast cancer. Phosphatidylinositol 3-kinase/AKT/ mammalian target of rapamycin(PI3K/AKT/mTOR) pathway is dysregulated in various cancer along with the breast cancer and is the major signaling pathway which controls the routine activity of our body including cell survival, cell division, etc. In this study, we found many pre-existing compounds against the mTOR protein and technically found SF1126 with the high-affinity properties along with the best binding energy score. Further research progress based on the structural similarity of SF1126 against drug database found that there are several more compound having property against mTOR. Best drug found in this research is ZINC85569445and concluded based on the ADME profile and BOILED-Egg plot prediction.
BOILED -Egg plot is a good informant for lead optimization and following analyses the virtual screened compound ZINC85569445 exhibits interesting pharmacokinetics. Virtual tested drugs are correctly placed in the white region of the egg which indicates that it constitutes preferable compound against mTOR protein. The binding free energy using post scoring method MMGBSA calculated for the evaluation of molecular docking, also concluded the virtual screened compound was having higher affinity with the receptor as compared to pre-exist drug.
Conclusively the lead compound is ZINC85569445 ideal for the study of pharmacophore profile. The compound shows the best affinity among all the pre-exit drugs, and inhibition property against the target mTOR as well as the study of pharmacophore mapping of the compound displayed considerable binding affinity in the active site of protein mTOR. This research encourages suitable opportunity for compound ZINC85569445 to prevent breast cancer.

Consent for publication
Not applicable.

Ethics approval and consent to participate
Not applicable.

Competing interests
The authors declare that they have no competing interests.