Screening and identification of the active components of Euphorbia thymifolia and identifying their potential targets implicated in breast cancer through Insilico approach.

Plant material collection and extract preparation by cold maceration method, GCMS analysis, Insilico molecular docking and molecular dynamics study.

The aerial parts of Euphorbia thymifolia L. were procured from the uncultivated areas of the Udupi District from April to June. The plant material was identified and authenticated by taxonomist Dr. K. Gopalakrishna Bhat, Professor of Botany (Rtd.) Poorna Prajna College, Udupi, The plant sample was deposited in the Herbarium of the Department of Pharmacognosy, Manipal College of Pharmaceutical Sciences. (Herbarium Voucher no: PP624) The aerial parts of Euphorbia thymifolia L. were collected, washed thoroughly and carefully with distilled water to remove all the soil and debris. The plant material was shade-dried at room temperature for 15 d until it was dry and crisp. After drying the shoots of E. thymifolia L., they were powdered and stored at 4°C.

The crude extract (10% w/v) 35 g in 350 ml was prepared by cold maceration as described by Abubakar et al using methanol procured from NICE Chemicals Private Limited (MI2037; B.no 103168) as the solvent for 72 h with intermittent shaking at room temperature. The extracts were then filtered through Whatman No. 1 filter paper. The clear extracts were concentrated using a rotary vacuum flash evaporator and freeze-dried by lyophilization, a process of drying the extract under controlled temperature and pressure using a lyophilizer. These lyophilized crude extracts were stored at 4°C until further use ( Abubakar and Haque 2020).

The composition of the methanolic extract of E.thymifolia was analyzed by GC-MS at Analytical Research & Metallurgical Laboratories Pvt. Ltd. Bangalore, India.

Molecular docking was performed using Schrodinger suite to select the best protein-ligand complex. Commercial Maestro software version 11.8 (OPLS3e force field) was utilized for all the simulation studies. Protein preparation wizard was used for preparation of protein using the panel review, modify, and refine modules where the side chains and residues were filled, and water molecules beyond 3 Å were removed. GLIDE panel was used for receptor-grid generation to create a grid and locate the receptor at the binding site for the ligands to be docked. The size of the grid box generated at the inbound ligand site for each protein was 10 × 10 × 10 Å ( Rathi et al. 2019).

The crystal structures of the proteins were retrieved from the Protein Data Bank ( Berman et al. 2000). ERK1 (Protein Data Bank ID: 4QTB, Crystallographic precision of the PDB: 1.40 Å) bound to a piperazine-phenyl-pyrimidine derivative ( Chaikuad et al. 2014); AKT (Protein Data Bank ID: 4EKL, Crystallographic precision of the PDB: 2.00 Å) with an ATP site inhibitor which is a piperazine- pyrimidine derivative ligand ( Lin et al. 2012); EGFR/HER2 (Protein Data Bank ID: 1XKK, Crystallographic precision of the PDB: 2.40 Å) with bound ligand Lapatinib is a tyrosine kinase inhibitor in clinical development for cancer ( Wood et al. 2004); ER (Protein Data Bank ID: 2IOG, Crystallographic precision of the PDB: 1.60 Å) with indole ligand ( Dykstra et al. 2007); The crystal structure of MELK (PDB ID: 5K00, Crystallographic precision of the PDB: 1.77 Å) with a ligand which is an amide derivative ( Chen et al. 2017); Crystal structure of PLK1 (PDB ID: 4J52, Crystallographic precision of the PDB: 2.30 Å) with pyrimidodiazepinone as ligand ( Nie et al. 2013), and Crystal structure of PTK6 (PDB ID: 5H2U,Crystallographic precision of the PDB: 2.24 Å) with ligand Dasatinib ( Thakur et al. 2017).

The structures of all 219 phytocompounds were either retrieved from PubChem or derived using Chem-Draw software. These 219 ligands were imported into maestro and “ LIGPREP” panel was utilized for ligand preparation ( Rathi et al. 2019). Docking was performed using the Maestro Glide Module. Initially, all ligands were docked in standard precision (SP) mode, followed by extra precision (XP) mode to obtain a ligand docking XP score ( Elokely and Doerksen 2013).

All XP docking files corresponding to each phytocompound and the seven proteins were then subjected to MM-GBSA free ligand energy calculation using the PRIME module to understand the binding energy of the target protein listed above with all 219 phytocompounds. ( Ignjatović et al. 2016).

To determine the protein ligand stability under the simulated physiological environment, the DESMOND panel of Maestro was used for the MD simulation assessment, which was executed on HER2 (1XKK), ERK1 (4QTB) with 3,6,9,12-tetraoxatetradecane-1,14-diyl dibenzoate (TTDB) (molecular formula: C24H30O8), and ER(2IOG) with succinic acid, 2-(dimethylamino) ethyl 4-isopropylphenyl ester (SADPE) (molecular formula: C17H25NO4) for 100 ns using the XP docking file. Three steps, namely the system builder, minimization, and MD simulation, were involved in this process. A predefined simple point charge model (SPC) was used to create an orthorhombic boundary for the XP docked complex of TTDB with 1XKK, 4QTB, and 2IOG with SADPE. The charges were neutralized, where three positive charges were neutralized by the addition of three chloride ions, and one negative charge was neutralized by one sodium ion. An isothermal–isobaric (NPT) ensemble with a constant temperature of 300 K and pressure of 1 bar was maintained throughout the simulation for 100ns. To determine the stability of the complex, the Root Mean Square Deviation (RMSD) was analyzed along with the protein–ligand contact timeline and covalent/non-covalent interactions ( Ivanova et al. 2018).

GC-MS analysis of the methanolic extract of E.thymifolia showed 23 prominent peaks, with each of the retention time peaks having several hits comprising a total of 245 hits, as shown in Supplemental Data File 1. Each of these hits was recognized by comparing their retention time peak, peak area (%), and height to those of the already known compounds identified by the National Institute of Standards and Technology (NIST) library. Of the 245 phytocompounds, duplicates were removed and 219 unique compounds were identified in the methanolic extract of E.thymifolia. The identified phytocompounds were organic acids, phenyl esters, hydrocarbons, saturated and unsaturated cyclic compounds, and fatty acids.

The results of XP molecular docking using GLIDE were as follows: All 219 ligands with each of the seven proteins were comprehensively analyzed with their docking scores and amino acid interactions. The ligands that showed good interaction with most of the proteins were 3,6,9,12-tetraoxatetradecane-1,14-diyl dibenzoate (TTDB), succinic acid, and 2-(dimethylamino) ethyl 4-isopropylphenyl ester (SADPE), as shown in Table 1. TTDB showed very good interaction with two proteins, namely 1XKK and 4QTB, with docking scores: MMGBSA ΔG binding energies of -8.627, -65.22 kcal/mol and -10.112 and -71.103 kcal/mol with two types of non–covalent interactions, respectively. Similarly, 2IOG with SADPE exhibited a docking score of -8.865 and MMGBSA ΔG binding energy of -45.94 kcal/mol with two types of non-covalent interactions.

The amino acid residue Lys745 of the 1XKK protein showed a strong hydrogen bond and Phe856 showed pi-pi stacking with TTDB. Lys 71 and Met 125 of the protein 4QTB interacted with the oxy- group of TTDB. The amino acid residue Lys531 of 2IOG showed a hydrogen bond, and Trp383 showed pi-cation interactions with the ligand SADPE, as illustrated in Table 2. Hydrogen bond formation between a ligand and a protein plays a crucial role in drug design. Furthermore, the XP file obtained was subjected to the Prime MM-GBSA method to evaluate the binding affinity more accurately. The calculated binding energies (MM-GBSA ΔG) for 1XKK and 4QTB with the TTDB complex were − -65.228 kcal/mol and -71.103 kcal/mol respectively, and that of 2IOG with SADPE was -45.945 kcal/mol.

Furthermore, to understand the interaction between the protein and ligand complex, a molecular dynamics study was performed for 100 ns. The RMSD values obtained were explained to obtain more clarity regarding the binding stability of the docked complexes.

The RMSD values of HER2 (1XKK) and TTDB were found to be around 4.8 Å, 4.0 Å respectively. The RMSD plot of both ligand and protein was stable after 20 ns, as shown in Figure 1. Later, the protein-ligand complex showed varied fluctuations until 50ns and then remained stable for the next 50ns. The molecular dynamics trajectory showed that amino acid residues such as Leu718, Ala 743, and Leu844 were vital before ligand (TTDB) stabilization up to 20 ns. Amino acid residues such as Val726, Leu844, and Phe856 established contacts with the ligand to be crucial hotspots for equilibrium, as shown in Figures 2B and 3.

To observe the protein-ligand interaction stability of TTDB with 4QTB, the RMSD values of the 4QTB with TTDB was found to be 2.2 Å and 2.4 Å respectively ( Figure 4). This complex exhibited the highest stability throughout the simulation without larger deviations. Amino acid residues, such as Ile 48, Ala 69, Ile73, Arg 84, and Asp 184, were significant in stabilizing the complex until 20 ns. Other amino acid residues, such as Tyr 53, Lys 71, Tyr 81, Met 125, Leu 173, Val 56, and Cys 183, showed strong contact with TTDB and contributed to their strong stabilization for 100 ns. TTDB showed a prominent interaction with ERK (4QTB) ( Figures 5B and 6 A & B).

The RMSD values of ER (2IOG) and SADPE were approximately 5.0 Å, 4.0 Å respectively. The RMSD plot of both ligands and proteins stabilized after 20 ns for some time ( Figure 7). Up to 20 ns, amino acid residues such as Thr347, Ala350, Leu 384, Asn532, Cys530, and Lys531 were important prior to ligand stabilization. Subsequently, the amino acid residues Asp351, Leu 387, Phe 404, Trp383, Leu525, and Tyr 526 showed strong interactions with the ligand, which contributed to stabilization ( Figures 8B and 9).