Exploring Natural Medicinal Plants for Novel Bcl-2 Inhibitors: In Silico Drug Design and Computational Interaction Mechanism for Anti-Cancer Activity.

The disturbance of typical cellular programmed cell death, known as apoptosis, constitutes a distinctive feature across all categories of malignancies. This maladjustment of apoptosis has the potential to result in diverse pathological states such as cancer, autoimmune illnesses, and neurodegenerative disorders. The regulation of apoptosis hinges upon proteins affiliated with the BcL-2 family, which possess the capacity to either foster or impede this progression. The exaggerated expression of anti-apoptotic proteins (Bcl-2, Bcl-xL, and Mcl-1) has been linked to the sustenance, proliferation, and advancement of tumors. Lately, there has been a surge of interest in investigating small compounds and peptides that possess the ability to bind to the BH3 binding pocket of these proteins, as they exhibit promising potential as agents against cancer. Initially, the primary emphasis in the development of anti-cancer agents targeting this protein family was centered on suppressing Bcl-2. However, the precise mechanisms of drugs specific to Bcl-2 and their impacts have not been fully clarified through computational approaches. By conducting


Introduction:
Cancer is the result of a disturbance in the normal programmed cell death activity of dividing, living cells.Programmed cell death, known as apoptosis, is a tightly regulated and highly conserved process that eliminates aged or damaged cell, (1) which is essential for cellular regulation and equilibrium, the extrinsic and intrinsic pathways (two well-regulated mechanisms) play a significant role in normal cellular conditions. (2)The extrinsic pathway relies on external factors like heat, radiation, and inadequate nourishment.Mitochondria controls the intrinsic pathway, (3) in which the Bcl-2 predominantly participates in this pathway and they have been extensively investigated as potential targets for diverse cancer treatments. (4)The Bcl-2 family encompasses both pro and anti-apoptotic proteins, with different Bcl proteins present in various regions of the human body. (5)There is association between anti-death proteins (Bcl-2, Bcl-XL, and Mcl-1) overexpression and tumor development, maintenance, and progression. (6,7)In all anti-death proteins the conserved Bcl-2 homology (BH) peptide domains-which involves 4 classes known as BH1, BH2, BH3, and BH4-are present. (8)BH3 is a 16-25 amino acid protein interaction motif predominantly found in pro-death members of the Bcl-2 protein family. (9)Peptides that bind to the BH3 domain attach to anti-death Bcl-2 proteins through a hydrophobic groove on their surface, promoting cellular demise. (10)The initial group of compounds investigated for Bcl-2 inhibition were peptides that imitated BH3-only proteins, exclusively targeting the BH3 region of anti-death proteins. (11)Peptidomimetic drugs were subsequently developed.Recently, clinical trials have been conducted on small molecule analogs of the amino acids involved in BH3 peptide binding interactions. (12)Drugs that bind to and hinder specific classes of Bcl-2 proteins have been also uncovered depending on fragment-based drug discovery and NMR-based structure prediction."ABT-199 is a selective inhibitor of Bcl-2". (13)It is currently undergoing Phase III clinical trials, exhibiting >4800fold greater selectivity (with a Ki of 0.01 nm) compared to other anti-apoptotic proteins such as Bcl-xL and Bcl-w, and it does not affect Mcl-1.Obatoclax is a broad-spectrum Bcl-2 antagonist (14) , with average IC50 values of 3 mM and 2.9 mM on Mcl-1 and Bcl-2, respectively (15) , it has successfully completed phase I and II clinical studies for various malignancies. (16)olecules such as ABT-737 and its orally active derivative ABT-263 demonstrate binding affinity towards Bcl-2 and Bcl-xL proteins, while not interacting with Mcl-1. (17)20) Due to their reduced adverse effects and undesirable reactions compared to synthetic medications, phytocompounds are the preferred choice.Consequently, it is crucial to test potent phytocompounds derived from medicinal sources against pathogenic bacterial proteins like ASPP2(Ank-SH3) to facilitate the development of novel and safe antimicrobial drugs. (21,22) cl-2 was selected as targets in this research due to its prevalence in most cancer types.Notably, Bcl-2 now possesses a well-defined threedimensional crystal structure (PDB ID: 4LVT).To identify natural lead-like phytocompounds against the Acinetobacter baumanni response regular "BfmR" enzyme and examine its molecular dynamics, a bioinformatics-driven hierarchy to search the IMPPAT database was employed. (23)Subsequently, molecular dynamics simulations (MDS) (24) were conducted to validate the binding affinity of the most promising phytocompound inhibitors within the BfmR active pocket.These inhibitors include "Norepinephrine, Australine, Calystegine B, 7,7 A-Diepialexine, and Alpha-Methylnoradrenaline".

1-Selection and preparation of ligands from the IMPPAT database
The Plant Phytocompounds Database (IMPPAT) is a repository for phytocompounds found in plants.Herbs from India gathered to find a ligand library for this study depending on the details in IMPPAT of 14,011 phycompounds. (25)As a result, the most widely utilized Lipinski's rule of five (Ro5), a drug-like filter, was used to screen the ligand library's potential drug-like compounds.There were 8,450 out of 14, 011 which were passed Ro5 (drug-like filter).After that, the threedimensional Structure Data File was downloaded (3-D SDF).The IMPPAT database provided the structure of these phytocompounds.Venetoclax and navitoclax are FDA-approved medications used as reference compounds in this study.

2-Preparation of protein
To conduct docking studies, the resolved crystal structure of the apoptosis regulator Bcl-2 (PDB ID: 4LVT) from the Protein Data Bank (PDB) website (www.rcsb.org)was obtained. (26)Prior to the docking analysis, the 3D crystal structure underwent necessary preparations, including the removal of water molecules and HETATM from the published structures.The energy minimization process employed the CHARMm force field, utilizing Discovery Studio Visualizer 2019.Furthermore, the active site of the pre-bound ligand molecules present in both structures was analyzed.All atoms and bonds within a <5Å radius of the selection zone was selected and the relevant amino acid residue information from the active site was collected.This information was then utilized to perform docking analysis on the corresponding binding pocket.

3-Molecular docking/ protocol
By performing molecular docking analysis, the inhibitory effects of all the investigated compounds against the 4LVT protein was evaluated.The protocol described in references 27 and 28 was followed.

A-Molecular docking protocol validation
To validate the docking protocol, we excluded the coordinates of the bound ligand inhibitor IXJ from the crystal complex of 4LVT and verified the bond orders.The protocol described in reference 28 was followed.

B-Pharmacoinformatics analysis
To ensure the comprehensive analysis of selected and newly designed compounds, we employed several pharmacoinformatics tools.The details described in reference 29.

C-Molecular Dynamic Simulation
To evaluate the stability of the protein backbone within the docking complex, we conducted a molecular dynamics simulation experiment using GROMACS version 2019.4 (28) .The Gromos54a7 all-atom force field (30) was employed for the simulation.The complexes with the lowest binding energies, as determined from the molecular docking analysis, were subjected to the simulation.For explicit solvation in the molecular dynamics (MD) simulations, we selected the SPC water model. (31)The dimensions of the rectangular box, ensuring a minimum separation of 1 nm between any atom and the box boundary to satisfy periodic boundary conditions, were determined to be 8.19 nm x 8.76 nm x 8.12 nm.The ProDRG webserver was used as source to obtain charge parameters for the ligand. (32)To balance the system's charge, Na + and Cl - atoms were introduced into the solution.The energy minimization process employed the steepest descent algorithm, with a maximum force (Fmax) threshold set at 1000 kJ/mol.nm.Subsequently, the system underwent equilibration at a temperature of 300 K and pressure of 1 bar, accomplished by conducting two consecutive 100 ps simulations using the canonical NVT and isobaric NPT ensembles, respectively.Throughout the equilibration phase, both the protein and docked ligands were independently restrained, and thermostat coupling was applied for the entire simulation.The MD simulations were run for 100 ns with stable temperature and pressure, utilizing a time step of 2 fs and a long-range interaction cut-off of 1 nm.Trajectory analysis was performed using GROMACS tools to extract relevant information from the simulations.

D-Binding free energy calculation
The binding free energy of each protein-ligand complex was determined using the MM/GBSA (Molecular Mechanics/Generalized Born Surface Area) approach based on the MD simulation trajectories.The gmx_MMPBSA package was employed for the calculations, utilizing snapshots extracted from the production MD trajectory at an interval of 100 ps from the 80-100 ns timeframe. (26,32)The calculation focused on evaluating the free energy (∆) associated with the formation of the ligand-receptor complex.The calculation done by equations described in references 28 and 33.In the MM/GBSA (Molecular Mechanics/Generalized Born Surface Area) approach various energy components-including "changes in the gas phase molecular mechanics energy (∆), solvation free energy (∆), and conformational entropy (-∆)" upon ligand binding-were considered to determine the binding free energy (∆).Fixed geometries before and after binding was obtained by cancelling out the bonded internal energy terms (∆, ∆, and ∆ℎ ).The nonbonded Van der Waals interaction energy (∆) and electrostatic interaction energy (∆) are also taken into account.The solvation free energy (∆) is composed of two components.The "polar contribution (∆)" is calculated using the "GB-OBC1" model, which considers the electrostatic solvation energy.To account for the nonpolar interactions between the solute and the continuum solvent, the "solvent-accessible surface area (SASA)" used for estimation the nonpolar contribution (∆). (

2-Pharmacoinformatic studies
Table 2 presents the results of the Drug Score and Drug Likeness model for the investigated compounds.It is crucial to acknowledge that compounds displaying zero or negative values should not be regarded as suitable candidates for drug-like properties.Among the analyzed compounds, Alpha-Methylnoradrenaline demonstrated the highest scores for both Drug Score and Drug Likeness, achieving values of 1.17.Subsequently, 7,7 A-Diepialexine and Norepinephrine obtained Drug Likeness scores of 1.15 and 1.07, respectively.These findings imply that these three compounds exhibit significant potential for further development as drug leads.Table 3 presents a summary of the descriptors for molecular properties of the compounds investigated, based on "Lipinski's Rules of Five".The calculations and theoretical framework for these descriptors were carried out following a previous study. (35) 3, none of the compounds studied violated any of "Lipinski's Rules of Five", indicating their potential as orally active drugs.Tables 4 and 5 provide a comprehensive overview of the ADMET properties, encompassing absorption, distribution, metabolism, excretion, and toxicity, for the compounds investigated.The ADMET profiles offer valuable insights into the potential of these compounds as drug candidates.The utilized database incorporates diverse ADMET characteristics that evaluate the capacity of the studied compounds to serve as drug leads.These characteristics include the penetration of the blood-brain barrier (BBB), human intestinal absorption (HIA), permeability across Caco-2 cells, CYP inhibitory promiscuity, AMES toxicity, carcinogenicity, and acute toxicity LD50 in rats.The outcomes of these evaluations are presented in Tables 4  and 5.As evidenced in Tables 4 and 5, all the examined compounds demonstrate BBB permeability and absorption in the human intestine (HIA furthermore, notable permeability for Caco-2 cells was observed, particularly for Isoliensinine, Liensinine, and Methylcorypalline.Notably, these compounds were found to have a favorable metabolic profile, as they were identified as non-substrates and non-inhibitors of CYP enzymes. (33)In terms of AMES toxicity, none of the studied compounds exhibited toxic effects.The carcinogenicity model also indicated the absence of carcinogenic properties in these compounds.Moreover, the rat acute toxicity LD50 values for all the compounds fell within the range of 2.10 to 3.19 mol/kg.The comprehensive information provided in Tables 4 and 5 strongly supports the potential of these investigated compounds as viable drug candidates, given their desirable ADMET properties.

3-Molecular dynamic simulation
To evaluate the stability of the docked complexes and examine the protein backbone stability within the docking complex, molecular dynamics simulations were conducted.Specifically, simulations were executed on the 4LVT protein in two distinct scenarios: in the absence of any ligand and in the presence of three particular ligands, namely "7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine".The simulations spanned a duration of 100 ns and employed an explicit solvation system.Additionally, to mimic physiological conditions, a physiological salt concentration was maintained throughout the simulations.

4-The root means square deviation (RMSD)
To assess the stability of the protein backbone in the presence and absence of ligands, the Root Mean Square Deviation (RMSD) values were calculated during the course of molecular dynamics simulations.As depicted in Figure 2, the RMSD values for the protein backbone demonstrated stabilization after 5 ns of simulation.
These values indicated comparable changes in the conformation of the protein backbone compared to the initial structures, with values below 0.50 nm.Notably, the apo-form of 4LVT exhibited the highest deviation in RMSD values, whereas the docked ligands, such as "7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine", displayed similar or lower RMSD values.This observation suggests that the docked ligands contribute to the stabilization of the 4LVT structure to some extent.The average RMSD values for the final 10 ns of the simulation were 0.34 ± 0.09 nm for 4LVT, 0.30 ± 0.11 nm for Alpha-Methylnoradrenaline within 4LVT, 0.29 ± 0.07 nm for 7,7 A-Diepialexine within 4LVT, and 0.26 ± 0.10 nm for Norepinephrine within 4LVT, respectively.

5-Root mean square fluctuation (RMSF)
The mobility and flexibility of protein residues during the simulation were analyzed by calculating the Root Mean Square Fluctuation (RMSF) values at the residue level.Figure 3 displays the RMSF values for the 4LVT protein both with and without ligands, including 7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine.As expected, the terminal residues of 4LVT and the docked ligands exhibited higher RMSF values, indicating their greater mobility.Specifically, in the apo-4LVT structure bound to 7,7 A-Diepialexine, residues 40 and 45 showed increased mobility.In the case of the apo-4LVT structure bound to Alpha-Methylnoradrenaline, residues 45, 55, and 70 exhibited higher flexibility.Additionally, the binding of Norepinephrine to 4LVT resulted in higher flexibility in residues 26-28.
These variations in RMSF values among different residues suggest differences in the binding characteristics of each ligand and their subsequent influence on protein structure and dynamics.Although Figure 3 shows similar RMSF values overall, specific residue-level changes can be observed upon binding of different ligands.

6-Radius of gyration (Rg)
To evaluate the overall impact of the ligands on the structure of 4LVT, the radius of gyration (Rg) values was computed for each complex, and the outcomes are depicted in Figure 4.The Rg values for the apo-4LVT and the 4LVT complexes with 7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine were determined to be (2.29,2.27, 2.26 , and 2.23) nm, respectively.Because Norepinephrine exhibits the lowest Rg value among the ligands this indicate that the protease structure when bound with becomes more compact when bound with Norepinephrine.

7-Hydrogen bonds (H-bonds) counts
The number of hydrogen bonds formed between main protein (4LVT: which is a potent lymphoid tyrosine phosphatase inhibitor in T cells) and the ligands can insights into the variations in the binding patterns of different ligands with the main protein, as well as the nature of their interactions.
Figure 5 illustrates the number of hydrogen bonds formed by the protease with "7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine" during the simulation.It is evident that 7,7 A-Diepialexine forms a greater number of hydrogen bonds compared to the other compounds.The average number of hydrogen bonds formed between 4LVT and the ligands "7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine" over the course of the 100 ns simulation, comprising 100,000 time points, were found to be 3.25, 3.15, and 2.99, respectively.Hydrogen bonds formed in an aqueous solution differs significantly for all four ligands in spite of similar should be binding energies exhibited by the docked complexes.This disparity can likely be attributed to the absence of solvent in the docking process, as the solvent environment can profoundly influence ligand binding.

8-Protein-ligand interaction energy/ analysis
Apart from molecular dynamics simulations, the MM/GBSA method was utilized to compute the binding free energies of the top three hit compounds, utilizing data obtained from the final 80 ns of the 100 ns MD simulation trajectories.The MM/GBSA analysis is a commonly employed technique for estimating interaction energies, which leverages molecular dynamics simulation trajectories to predict the binding free energies (ΔGbind) of protein-ligand systems.The calculated ΔGbind values for the ligands remained negative, indicating favorable binding, as demonstrated in Table 6.
Among the investigated compounds "7,7 A-Diepialexine" exhibited a consistently negative mean value of -33.90 ± 7.16 for the binding free energy, surpassing the other compounds.This suggests that nonpolar interaction energy plays a significant role in the binding process, emphasizing the importance of hydrophobic interactions between "7,7 A-Diepialexine" and 4LVT.The negative electrostatic energy indicates the presence of favorable intermolecular hydrogen bonds.It is noteworthy that in all three studied systems (4LVT-"7,7-A-Diepialexine", 4LVT-"Alpha-Methylnoradrenaline", and 4LVT-"Norepinephrine"), the polar solvation energy opposes the binding, but this is counterbalanced by favorable van der Waals energy, electrostatics energy, and nonpolar solvation energy.

Conclusion
Through molecular docking, significant binding potential of "Norepinephrine, Australine, Calystegine B, 7,7 A-Diepialexine, and Alpha-Methylnoradrenaline" to the 4LVT protein compared to FDA approved drugs was observed.Molecular docking simulations further revealed strong binding of "7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine" which stabilize the structure of 4LVT by forming stable complexes.These findings suggest that 4LVT can serve as a potential target for novel drug development using "7,7 A-Diepialexine, Alpha-Methylnoradrenaline, and Norepinephrine" as promising lead phytocompounds.Experimental validation through in vitro and in vivo studies can be conducted to explore their potential for drug discovery against A. baumannii.Lipinski's rule was established based on five rules to compute the ability of the compound to act as an orally active drug.According to Lipinski's rule, the orally active drug must have no more than one violation of the following standards: (i) octanol/water partition coefficient (logP), which measured the lipophilicity of a molecule must be not greater than five; (ii) a molecular weight (MW) less than 500 Da; (iii) not more than five hydrogen bond donors (nOHNH); (iv) not more than 10 hydrogen bond acceptors (nON); and (v) topological polar surface area (TPSA) below the limit of 160 Å. (36) So, Compounds with higher weights are less likely to be absorbed and therefore to ever reach the place of action. (37,38) f we apply the molecular weight pillar on our compounds, we can say all studied compounds are obey Lipinski's rule.

Table 6 . MMPBSA energy (ΔE MMPBSA ) studies
The author(s) declare no conflict of interest, financial or otherwise.