Pharmacokinetics of some newly synthesized 1, 5- benzothiazepine scaffolds: A molecular docking and molecular dynamics simulation approach

2.1 Synthesis of chalcones (3a-n)

Equimolar quantities 4-hydroxyacetophenone (0.01 mol) and substituted aryl aldehydes (0.01 mol) were put in ethanol (30 mL) and agitated at room temperature. With constant stirring, aqueous solution of KOH (15 mL, 40 %) was put to the reaction combination. The reaction combination was retained overnight and transferred onto ice water and diluted with dilute HCl. The solid acquired was strained off and recrystallized from ethanol to obtain chalcones.

2.2 Synthesis of 1,5- benzothiazepine derivatives (5a-n)

To a minimum quantity of ethanol, substituted chalcones 1 (0.01 mol) were dissolved. To this, 2- aminothiophenol 2 (0.01 mol) and a pinch of CuO catalyst were put in and the resulting reaction combination was refluxed for 3 h at 60 °C. Afterwards, the reaction mixture was acidified by the addition of glacial acetic acid (5–6 drops) and reflux was continued for further 2–3 h. The achievement of the reaction was examined by TLC [Benzene: Ethylacetate; 9:1 v/v]. The reaction combination was permitted to cool at room temperature and the content was transferred against ice water. The mixture was filtered off and solid obtained was purified by column chromatography to afford the title product [(Supplementary Information: Appendix III (5a-n)].

2.3 In silico pharmacokinetic analysis of 1,5- benzothiazepine derivatives

Molinspiration server (https://www.molinspiration.com) was employed for analyzing the drug likeliness and molecular descriptors of synthesized 1,5- benzothiazepine derivatives. As per Lipinski’s rules of five, a drug like molecule must have logP ≤ 5, molecular weight ≤ 500, number of hydrogen bond acceptors ≤ 10, and the number of hydrogen bond donors ≤ 5. And, violation of any of the above rule may affect the oral bioavailability. $$\mathrm{P}\mathrm{e}\mathrm{r}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{o}\mathrm{f}\mathrm{a}\mathrm{b}\mathrm{s}\mathrm{o}\mathrm{r}\mathrm{p}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}(\%\mathrm{A}\mathrm{B}\mathrm{S})=109-0.345\times \mathrm{T}\mathrm{P}\mathrm{S}\mathrm{A}$$

where, TPSA = Total Polar Surface Area.

The prediction of the bioactivity score of drugs was performed by calculating vital molecular descriptors such as logP, polar surface area, number of hydrogen bond donors and acceptors. Further, ADMET profiling (Absorption, Distribution, Metabolism, Excretion and the Toxicity) of derivatives was performed using admetSAR (Bhardwaj et al., 2019; Khan et al., 2015). admetSAR is a web-base tool containing manually curated data of various known chemical compounds and serve as a controller for determining ADMET properties of novel input compounds sharing structural and functional similarity.

2.4 Potential therapeutic target identification

Systematic identification of potential therapeutic targets for 1,5- benzothiazepine derivatives was performed by using in silico consensus inverse docking (ACID) server. ACID is a consensus platform is capable of figuring the binding energies of docked complexes from AutoDock Vina, LEDOCK, PLANTS, and PSOVin (Wang et al., 2019). In order to expand the accuracy of the docked pose and binding mode prediction, a system level drug repurposing task was performed by integrating Molecular Mechanics/Poisson–Boltzmann Surface Area (MM/PBSA) and X-SCORE for concluding binding energy calculation (Singh et al., 2017). This includes following steps:

1. Selection of test dataset containing 16,151 protein–ligand entries from PDB-bind database;

2. Preparation of 3D structure of proteins by removal of water molecules using PDB2PQR 1.6 21 program;

3. Automated active site prediction of target proteins followed by the generation of hundred conformations for every ligand versus target protein’s active site;

4. Calculation of binding free energy (ΔG_bind) via MM/PBSA and X-score methods, under below equation:

2.5 Molecular dynamics (MD) simulation

MD simulation of the stated docked complexes was supported by the usage of GROMACS package. In order to research the conformational balance, MD simulations have been done by following the same old tactics. For the era of ligand topologies and random human error elimination, the ACPYPE device was used. By employing the single factor price (SPC) water model and the inclusion of water molecules, a solution was created. Ordered water molecules serve as a guide for protein–ligand binding affinity by bridging the interaction between the two (Khan et al., 2018; Khan et al., 2018). Complete neutralization was carried out by adding of counter ions (Na⁺ and Cl^-) at 0.15 M of biological salt awareness. A particle-mesh-based Ewald (PME) method was used to calculate the electrostatic force and strength, which helps to assign particular periodic boundary conditions (Gupta et al., 2020). Lennard-Jones interactions cut-off was set to 10 to limit the motion of all covalent bonds pertaining to hydrogen atoms. To investigate the MD simulations' paths toward electricity minimization, a six-step approach was used (a) using the Broyden-Fletcher-Goldfarb-Shanno (LBFGS) set of principles, solvent molecule energy minimization prior to the device as a whole, (b) To reach the equilibrium state, non-hydrogen solute atoms were limited to 300 K temperature and 1 bar for 40 ns, (c) Initial simulations were conducted at controlled temperatures and pressures using Berendsen thermostats and a set of barostat rules, (d) The study of the root-mean-square (RMSD) deviation is aided by the initial trajectory information, (e) root-mean-square fluctuations (RMSF), and (f) radius of gyration (Rg) for protein–ligand complexes (Anwer et al., 2020).

3.1 Synthesis of CuO catalyst, chalcones and 1,5-benzothiazepines

Our research aim was to develop an efficient methodology for the production of 1,5-benzothiazepine derivatives, which is a seven membered heterocyclic compound that contains N- and S-atoms.

Firstly, the desired heterocyclic CuO catalyst was prepared by using a reported method (Luna et al., 2015). The synthesized catalyst was categorized by SEM and XRD techniques. The SEM images and XRD pattern are presented in Figure S-1 (Supplementary Information).

Through Claisen-Schmidt condensation reaction, the chalcones were prepared by reacting p-hydroxyacetophenone with various substituted aldehydes in a basic medium (KOH). The goal compounds (5a-n) were produced by Michael addition of 2-aminothiophenol to chalcones in acidic medium in the presence of CuO catalyst (Scheme S-1 and Appendix- I, II and III; Supplementary Information).

3.2 In silico pharmacokinetic analysis of 1,5- benzothiazepine derivatives

Pharmacokinetic parameters such as CLogP, lipophilicity, molecular weight and topological polar surface area (TPSA) of the newly synthesized 1, 5-benzothiazepine derivatives were computed using Molinspiration server and are summarized in Table 1.

3.3 Molecular dynamic (MD) simulation

Among seven selected 1,5- benzothiazepine derivatives (5b, 5d, 5f, 5 h, 5j, 5 k, and 5 l), consensus inverse docking against PDB-binding database identified potential therapeutic targets using structure-based screening through ACID server (Table 3).

4.1 Analysis of in silico pharmacokinetic properties

The sum of fragment-based contribution and correlation factors are termed as CLogP = octanol/water partition coefficient. Based on Lipinski’s rule of five, compounds having molecular weight ranging between (292–470 i.e., > 500) were selected as low molecular weight compounds and can be easily diffused, transported and absorbed in comparison with heavy compounds. If a chemical becomes too bulky, it can interfere with the therapeutic effects of drugs. By screening molecules with hydrogen bond acceptors and donors fewer than 10 and 5, respectively, it was possible to prioritise the compounds. The permeability of compounds was assessed by computing logP value less than 5. This allows easy penetration of drugs through biomembranes. The computation of topological polar surface area (TPSA) based on O^- and N^- centered polar fragments assists in characterization of drug and intestinal absorption, bioavailability, Caco-2 penetrability and blood–brain barrier diffusion. Both, the degree of lipophilicity and TPSA indicate strong interaction of drug with membranes and hydrogen bonding potential of the compound. Higher quantity of rotatable bonds raises the flexibility of molecule and resultantly adaptability for effective interactions within binding pocket. Whereas, admetSAR enables the computation of absorption, distribution, metabolism, excretion and toxicity (ADMET) properties. For entire derivatives, blood–brain barrier (BBB) permeation, HIA (human intestinal absorption), Caco-2 cell permeability and AMES test (a type of biological assay for assessing the mutagenic potential of chemical compounds) were considered and shortened in Table 2.

For efficient drug absorption and allowance in the liver, concentration of several isoforms of cytochrome P450 super family (i.e., CYP1A2, CYP2A6, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4) were estimated. This identifies that ROS acts as a substrate for P-glycoprotein that supports the efflux of drug before metabolism leads to therapeutic drug failure and metabolic clearance. Based on predicted values of admetSAR, seven ligands derived from 1,5- benzothiazepine (5b, 5d, 5f, 5 h, 5j, 5 k, and 5 l) were able to penetrate blood–brain barrier permeation, Caco-2 cell and human intestine and showed no toxicity or mutagenic effect.

4.2 Inverse molecular docking

Comparative analysis revealed 4-hydroxyphenylpyruvate dioxygenase (Uniprot ID: P32754), dimethylglycine dehydrogenase (Uniprot ID: Q9AGP8) and folate receptor beta (Uniprot ID: P14207) as major targets for the selected 1,5- benzothiazepine derivatives. In addition to the above stated targets, muscarinic acetylcholine receptor M4; leukotriene A-4 hydrolase; cytochrome c oxidase subunit 2; progesterone receptor; histamine N-methyltransferase were also screened as potential targets. In case of 4-hydroxyphenylpyruvate dioxygenase (HPPD), 5j and 5 k served as potential inhibitors with binding energies of −11.76 kcal/mol and −11.56 kcal/mol respectively. For dimethylglycine dehydrogenase (DMGDH), 5b and 5d showed inhibiting activity with binding energies of −9.22 kcal/mol and −10.54 kcal/mol respectively. For folate receptor beta, 5f and 5j computed binding energies of −10.05 kcal/mol and −10.32 kcal/mol respectively.

4.3 Description of main targets for the selected 1,5- benzothiazepine derivatives

4-hydroxyphenylpyruvate dioxygenase (HPPD): It is a Fe (II)-containing oxygenase enzyme involved in catabolic transformation of 4-hydroxyphenylpyruvate into homogentisate in tyrosine metabolism. Literature citation revealed linkage of HPPD with metabolic disorders such as alkaptonuria and Type III tyrosinemia (Yang, 2003). Being involved in industrial production of necessary co-factors in plants related to herbicide development, HPPD inhibitors have equal importance in agribiotechnology. High level of tyrosine concentration in blood due to lower HPPD enzyme concentration in human intestine results in lenient mental retardation at birth and degradation in vision (Hühn et al., 1998). Therefore, HPPD is considered as a potential goal for drug and pesticide finding (Lin et al., 2019). Docking analysis revealed that Phe424, Asn423, Tyr456, Glu394, Asn282, Gln307 and Ser267 play a significant part in the binding of C-3 position of hydroxyl group in the benzene ring and sulfur group of both substrate residues (Figure S-3 a,b; Supplementary Information).

Dimethylglycine dehydrogenase (DMGDH): It is a mitochondrial matrix flavoprotein, which is active in choline and 1-carbon metabolism. It is involved in catalytic demethylation of dimethylglycine to form sarcosine that bounds to FAD cofactor in oxidative phosphorylation machinery. Higher level of DMGDH in blood stream is accountable for metabolic disorders related to choline metabolism (Binzak et al., 2001). Docking studies revealed that amino acid residues in active ATP binding sites (GLU 915, ARG 927, LEU 868, GLY 920, GLU 883, LYS 885, PHE 916, PHE 919, LYS 1053 and ASN 921) are hydrogen bonded to methoxy group at C3 and C9 position of 1,5- benzothiazepine derivatives. The presence of van der Waals interactions with Met 68, Val 69 and Leu 90 with C4 group of benzene ring are important stuctural requirenent for anti-toxicity, which was also found during the analysis (Figure S-3c,d; Supplementary Information).

Folate receptor beta (FR): It is a glycosylphosphatidyinositol-linked protein capturing ligand from the extracellular matrix and carries them inside the cell through endosomal pathway (Yi, 2016). It binds to folate and reduces folic acid derivatives mediating 5-methyltetrahydrofolate inside the cell. It is considered as a potential tumor biomarker. High expression level of folate receptor (FR) occurs in human malignancies, ovarian, endometrial and colon cancer. The FR is an emerging therapeutic target for the cure of chronic inflammatory diseases and cancer (Chandrupatla et al., 2019). The results of docking showed that the major H-bonding interaction with benzene ring (C-9th position of –OH group and the phenyl ring D) could associate with the (O-atom) key chain of Tyr88, Phe86, Asn124, Gln165 (Figure S-3 e,f; Supplementary Information). Tyr88 and Phe86 were found to be the main active residues accountable for hindrance of folate receptor beta.

The firmness of protein–ligand complexes was further examined by molecular dynamics simulations. The backbone of initial structure of the protein target was utilized for the generation of root mean square deviation (RMSD) graphs with respected to time at 40 ns with an average RMSD computed as 0.334 ± 0.05 nm (Ahmed et al., 2020). In case of 4-hydroxyphenylpyruvate dioxygenase Fig. 1 (a,b,c) and dimethylglycine dehydrogenase Fig. 2 (a,b,c) related complexes, RMSD values gained till 10 ns and conjunction was detected from 40 ns but slight variations stayed throughout. The overlaid graphical demonstration of time-dependent RMSD of protein–ligand complexes are represented in Fig. 1(a), Fig. 2(a) and 3(a). The dictionary of secondary structure of proteins (DSSP) suggests that active sites lies in the coil region and the majority of binding sites shaped β-helix structure. Residue-based root mean square fluctuations (RMSF) study of protein–ligand docked complexes was accomplished to know the flexibility of each residue and represented in Fig. 1(c), Fig. 2(c) and 3(c). RMSF values for all the docked complexes had large fluctuations (0.33–1.67 nm) for initial 20 residues in each case due to unavailability of structural information of the target protein. Further, lesser fluctuations at binding and active site indicate rigidity and intactness of the binding cavity. gmx_gyrate calculates Rg values indicating solidity and structural alterations of the docked complexes. Additionally, it calculates the atomic mass that corresponds to the complexe’s centres of mass. Please refer, Fig. 1(b), Fig. 2(b) and 3(b). Average Rg values of 4-hydroxyphenylpyruvate dioxygenase, dimethylglycine dehydrogenase and folate receptor beta ranged between (2.36–2.66 nm), (2.29–2.89 nm) and (2.44–2.56 nm) respectively with no fluctuations after 25000 ps. Further, Rg values correlate with RMSD values of backbone Cα atoms validating the firmness of the prioritized protein ligand complexes. This justifies the suitability of the prioritized 1,5- benzothiazepine derivatives to be further explored as potential drugs against various human disorders.

Close

Fig. 1

MD simulation of 4-hydroxyphenylpyruvate dioxygenase-5j and 4-hydroxyphenylpyruvate dioxygenase-5 k: (a) RMSD; (b) Rg; (c) RMSF.

Export to PPT

Close

Fig. 2

MD simulation of Dimethylglycine dehydrogenase-5b and dimethylglycine dehydrogenase-5d: (a) RMSD; (b) Rg; (c) RMSF.

Export to PPT

Close

Fig. 3

MD simulation of folate receptor beta −5f and folate receptor beta −5j: (a) RMSD; (b) Rg; (c) RMSF.

Export to PPT