Exploring the efficacy of some biologically active compounds as anti-hypertensive drugs: an insightful evaluation through DFT, molecular docking and molecular dynamics simulations

Abstract

Among different anti-hypertensive drugs, calcium channel blockers and human angiotensin-converting enzyme (ACE) inhibitors are the two main types. Herein, we took 25 biologically active ligands with potent anti-hypertensive activities and performed molecular docking studies with the human ACE receptor (PDB ID 1O8A) and human leukocyte antigens (HLA) complex, human voltage-dependent calcium channel alpha1 subunit (PDB ID 3LV3). Beforehand, we had performed density functional theory (DFT) studies to find out their structure–property relationships. In-silico ADMET studies were conducted, and we found that all 25 ligands follow Lipinski’s Rule of 5, which confirms their oral bioavailability and high gastrointestinal absorption as a drug. Finally, molecular dynamics (MD) simulation studies were performed for the two top-scored drugs for 100 ns which reveal that a strong influence of the ligand (flunarizine) is there over the respective proteins.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40203-024-00291-4.

Introduction

Hypertension is considered to be one of the leading risk factors for developing various cardiovascular conditions, and it significantly contributes to morbidity and mortality worldwide (Katouah and Gaffer 2019). Several critical diseases e.g., heart attack, renal failure, and high blood pressure all are associated with hypertension. As per the World Health Organization (WHO), almost 1/3rd of the adults suffering from hypertension died mainly because of angio-cardiopathy death and renal failure (Fadahunsi et al. 2022; Gayathiri et al. 2024). Anti-hypertensive drugs can be classified mainly into 5 types- a) angiotensin 2 receptor antagonists (AT2R), b) angiotensin-converting enzyme (ACE) inhibitors, c) beta blockers, d) calcium channel blockers, and e) loop diuretics. Apart from the risk of angio-cardiopathy death, the risk of developing high blood pressure is also associated with hypertension (Ahmed et al. 2018). As per the report of WHO, the risk of mortality between the age group of 30–79 years due to high blood pressure has increased by 33%. Very recently, Oktaviani et al. (Oktaviani et al. 2023) identified bisoprolol as one of the key medicines for treating hypertension. The authors have also shown that the addition of calcium channel blocker antagonists with bisoprolol can help a lot in the treatment of hypertension. In the year 2013, a joint group of scientists from the American College of Cardiology Foundation (ACCF) and the American Heart Association (AHC) reported hypertension as a major concern of heart failure (Yancy et al. 2013). Globally, there are many therapeutic classes of anti-hypertensive agents i.e., sulfonamide, 1,4-dihydropyridine (DHP), xanthone, etc. Calcium channels are normally used as building blocks and regulate many metabolic regulations. 1,4-dihydropyridine (DHP) type of compounds belong to the class of calcium channel blockers and they are widely used as anti-hypertensive agents (Saddala et al. 2017). Recently, Goshain et al. (2019) have reported the antihypertensive and toxicity of some synthesized xanthone derivatives based on ADME (Absorption, Distribution, Metabolism, and Excretion) properties and molecular docking studies. Noteworthy, xanthone derivatives are a class of thiazide loop diuretics.

Calcium channel blockers can be classified into two main therapeutic classes (a) non dihydropyridines and (b) dihydropyridines (DHP_S). Among non-dihydropyridines and dihydropyridine derivatives, dihydropyridines (DHPs) can be classified as L-type VGCC (Voltage Graded Calcium Channel) blockers. Both of the two classes of molecules are active on cardio-vascular cells (Koçak Aslan et al. 1307; Malek et al. 2024). Apart from angio-cardiopathy death, dihydropyridines can cross blood–brain barrier (BBB) permeability and can also be used for the treatment of several viral diseases e.g., Hepatitis C, HIV, etc. (Smati et al. 2024) Nimodipine belongs to the member of L-type calcium channel blockers. Being a lipophilic compound, it shows BBB permeability, and thus nimodipine can also be used as a drug for neurological diseases (Türkeş et al. 2021). Aside from the class of calcium channel blockers, angiotensin-converting enzyme (ACE) inhibitors are treated as crucial in the pathway of treating hypertension as they inhibit the conversion of angiotensin 1 to angiotensin 2 (Attique et al. 2019; Passos-Silva et al. 2015; Bader 2010). In the case of the treatment of hypertension, ACE is a key enzyme that regulates blood pressure in the human body. A group of scientists did an anti-hypertensive property study of certain egg white (EW) derivative peptides in order to find which of them possess in-vitro activity and antioxidant properties (FitzGerald et al. 2004; Miguel et al. 2005). Besides calcium channel blockers and angiotensin-converting enzyme (ACE) inhibitors, loop diuretics and angiotensin 2 receptor antagonists belong to another two classes of antihypertensive agents. Among several pharmacological agents used so far, loop diuretics belong to another class of pharmacological agents, which can be used for the treatment of hypertension. Musini et al. (2015) did a systematic review of loop diuretics intending to show the effect of loop diuretics in the treatment of cardiovascular disease. Besides these, chronic heart failure belongs to another class of hypertensive disease. It has been seen that despite using ACE inhibitors and beta blockers, the risk of worsening heart attack remains in play (Huo et al. 2019; Xiang et al. 2024). To overcome this difficulty, one needs to use angiotensin 2 receptor antagonists (AT2R) as medication. In recent years, several researches were carried out to find the use of AT2R. Sulfonamides belong to the sulphur related group of antibiotics, which can be used to treat several bacterial infections as well as to treat hypertension (Zhu et al. 2024). Arya and his co-workers performed in-silico investigations on different chemical constituents of the plant Clerodendrum Colebrookianum and found acetoside with the highest Glide docking score (− 12.97) compared to the co-crystallised inhibitor lisinopril (− 9.64) (Drug and Rock 2019). Durdagi and his co-workers utilized both computational and experimental approaches for designing novel herg-neutral anti-hypertensive oxazolone and imidazoline derivatives. The authors took antagonist bound AT1 receptor (PDB ID 4Y4Y) and several commercial oxazolone and imidazoline molecules for their study (Durdagi et al. 2018).

In this particular work, we selected the specific 25 compounds, as these compounds were guided by their clinical relevance and diversity in different anti-hypertensive therapeutics. As we already mentioned above using various literature sources, these compounds include different widely prescribed anti-hypertensive agents e.g., ACE (Angiotensin Converting Enzyme) inhibitors, ARB_S (Angiotensin Receptor Blockers), and calcium channel blockers (CCB_S), angiotensin 2 receptor antagonists (AT2R) and loop diuretics. Among these widely prescribed anti-hypertensive agents, ACE inhibitors and CCBs are two important types, as they address different underlying mechanisms of high blood pressure across various co-morbid patients. Several researches were done to find out the biological activity of these two types of anti-hypertensive agents. Recently, Yin et al. did various experimental and molecular modeling studies for investigating the kinetics and reactivities of the calcium channel blockers, and they found CCBs have different reactivity with halogen oxidants and photochemical transients (Yin et al. 2024). To find out the anti-diarrheal effect of a natural alkaloid piperine through interaction with inflammation-induced enzymes (cyclooxygenase 1, cyclooxygenase 2, L-type calcium channel blocker, etc.) Afroz et al. took two anti-hypertensive agents loperamide and bismuth subsalicylate as controls. By means of in-silico and in-vivo studies, the authors found piperine as a potent drug for anti-diarrheal agents (Afroz et al. 2024). Apart from CCBs, ACE inhibitors belong to another class of anti-hypertension therapeutics. Very recently, J. Li. and his co-workers with the help of an LC–MS characterization, mechanism, and structure–activity relationship (SAR) study find anti-hypertensive effects of different ACE inhibitory peptides after screening from sulfur hydrolytase (Li et al. 2024). In our study among the selected 25 compounds (most of the compounds mainly belonged to CCBs and ACE inhibitor types), we identified flunarizine as the most potent inhibitor in anti-hypertensive treatment, as it selectively blocks calcium entry in a person (calcium channel blockers) with additional neuroprotective properties due to its BBB permeability nature. In order to justify the statement, we did a molecular docking study of 25 ligands with the two receptors (PDB ID 1O8A and PDB ID 3LV3) and we found 1O8A-flunarizine and 3LV3-flunarizine complexes as the two best-docked complexes. Before doing molecular docking, we had performed geometry optimizations on the 25 ligands by using density functional theory (DFT) based electronic structure calculations to reproduce several electronic properties such as energies of highest occupied and lowest un-occupies molecular orbitals (E_HOMO, E_LUMO)_, electronegativity, dipole moment, etc. of the compounds, which correlate well with the docking result. Finally, we did an MD simulation study of the two best-docked compounds i.e., 1O8A-flunarizine and 3LV3-flunarizine for 100 ns. After completing the MD simulation study, we plotted RMSD (Root Mean Square Deviation), RMSF (Root Mean Square Fluctuation), R_g (Radius of gyration), and SASA (Solvent Accessible Surface Area) for the two best-docked complexes, which indicate the strong influence of the ligands concerning the two protein moieties (PDB ID 1O8A and PDB ID 3LV3).

Methodology

Protein preparation of docking

As already mentioned, we took two protein residues in order to find out the better potent inhibitor, which can behave like an anti-hypertensive drug. For this purpose, we downloaded the protein residues i.e., human ACE inhibitor (PDB ID 1O8A) and L-type calcium channel blocker (PDB ID 3LV3) from the RCSB protein data bank (Rose et al. 2017) (www.rcsb.org) in PDB (Protein Data Bank) format. Then, with the help of Auto Dock 4.2 (Morris et al. 2009) and MGL Tools, at first, water and other heteroatoms were deleted from the protein moiety. After that, by adding polar hydrogens and Kohlmann charges, the protein moiety got saved into PDBQT format and got ready for docking. The 3D structures of the two proteins have been represented in Fig. 1.

Fig. 1.

3D structures of a 1O8A and b 3LV3 visualized using Python molecular viewer

Ligand preparation for docking

All the 25 ligands were optimized within DFT using the B3LYP/6-31G level of theory. The ligands were then converted into PDB format and after that with the help of Auto Dock tools, the ligands were saved into PDBQT format.

Molecular docking study

After the preparation of the protein and the ligands, both the receptor protein molecules and the ligands were enclosed in a grid box with 114 × 108 × 120 Ǻ³ grid dimension and with a spacing value of 0.675 Ǻ. Lamarckian GA and other default parameters were used to carry out the docking procedure by using Auto Dock Vina (http://vina.scripps.edu/) (Trott and Olson 2010) package. By using Auto Dock Vina tool, preliminary we ran the molecular docking study with a default global exhaustiveness of 8 and 9 default no of GA (Genetic Algorithm) runs. In order to fit the ligand within the cavity of the receptors, the grid box was set at center_x = 40.59 Ǻ, center_y = 37.373 Ǻ, and center_z = 43.447 Ǻ points for the docking study of 1O8A receptor and the 25 ligands- the dimension of the grid box set at 114 × 108 × 120 Ǻ³. In the case of the 3LV3 receptor with the 25 ligands, the centers of the x, y and z coordinates were set as center_x = 38.578 Ǻ, center_y = 9.803 Ǻ and center_z = 19.003 Ǻ, with the same grid dimensions. The respective docking scores of 25 ligands with the two proteins have been represented in Table 1.

Table 1.

Docking scores of 25 ligands (the lowest binding energy) with the human ACE inhibitor (PDB ID 1O8A) and L-type calcium channel blocker antagonists (PDB ID 3LV3)

Compound name (Ligand)	PubChem CID	Docking score with (1O8A) (kcal/mol)	Docking score with (3LV3) (kcal/mol)
Alacepril	71992	− 7.3	− 5.1
Amlodipine	2162	− 6.0	− 6.6
Benazepril	5362124	− 8.2	− 6.8
Cilazapril	56330	− 8.6	− 5.5
Cilnidipine	5282138	− 8.1	− 5.4
Clevidipine	153994	− 7.6	− 6.9
Diltiazem	39186	− 7.4	− 7.1
Enalapril	5388962	− 7.5	− 7.2
Flunarizine	941361	− 10.3	− 10.2
Fosinopril	55891	− 6.6	− 5.6
Imidapril	5464343	− 7.0	− 7.4
Lercanidipine	65866	− 9.1	− 8.6
Levamlodipine	9822750	− 6.3	− 5.0
Moexipril	91270	− 6.7	− 7.3
Nicardipine	4474	− 8.7	− 7.8
Nifedipine	4485	− 5.2	− 5.1
Nisoldipine	4499	− 8.1	− 5.1
Perindopril	107807	− 6.2	− 6.0
Barnidipine	443869	− 7.9	− 8.6
Benidipine	656668	− 9.2	− 9.8
Quinapril	54892	− 6.8	− 6.9
Ramipril	5362129	− 8.8	− 6.5
Verapamil	2520	− 5.7	− 4.5
Zofenopril	92400	− 7.9	− 8.1
Nimodipine	4497	− 6.1	− 5.1

Conceptual DFT study of the ligands

Before the molecular docking procedure, we had optimized all the 25 ligands by using Gaussian 16 package by using 6-31G basis set and B3LYP (Becke, 3-parameter, Lee–Yang–Parr) functional in order to estimate their molecular properties such as E_HOMO, E_LUMO, etc. which correlate with their several global reactivity descriptors are listed below (Vijayaraj et al. 2009; Chakraborty and Chattaraj 2021; Proft et al. 2006).

MD Simulation study of the 1O8A-flunarizine and 3LV3-flunarizine complexes

In this work, we had performed Molecular Dynamics (MD) simulation studies of the best two docked compounds (each of one protein) by using GROMACS software package (Version 2022.4) (Berendsen et al. 1995).In order to do the process with the most stable conformer (having the highest binding energy) obtained from the molecular docking studies, we had used the CHARMM36-jul2021 force field (Huang et al. 2017) by TIP3P solvation model (Lee et al. 2014). By using the CHARMM General Force Field server, we had generated the topological and several parameter files for the ligand i.e., for flunarizine. For maintaining periodic boundary condition, both 1O8A-flunarizine and 3LV3-flunarizine complexes were set at least 1 nm distance from the cubic box. In the case of 1O8A-flunarizine complex 13 Na⁺ ions and for 3LV3-flunarizine complex 9 Na⁺ ions were added to neutralize both systems. Energy minimization of the systems was carried out with the help of steepest descent algorithm. In order to carry out the energy minimization of the overall protein–ligand systems using steepest descent algorithm 50,000 steps were performed until the F_max converged to < 1000 kJ/mol. After that, NVT ensemble was used for the equilibration of both the two systems for 100 ps at 300 K. NPT ensemble was then used for both the two systems for 100 ps at 300 K. For both NVT and NPT ensembles 2 fs time step were taken. In case of the 1O8A-fluranizine system, electrostatic and van der Waals cut off were kept at 2.0 nm for both NPT and NVT equilibration, whereas for 3LV3-flunarizine system both electrostatic and van der Waals cut off were kept at 1.2 nm for both NPT and NVT equilibration. In order to calculate the long-range interactive forces smooth particle mesh Ewald (PME) method (Abraham and Gready 2011) was used. With the same electrostatic and van der Waals cut off, the NVT and NPT ensembles were finally subjected to undergo molecular dynamics simulation (MD simulation) for 100 ns of each protein.

The structural trajectories of both 1O8A-fluranizine and 3LV3-flunarizine complexes were calculated using trjconv tool. With the help of Xmgrace plotting tool, RMSD, RMSF, R_g and SASA (Solvent Accessible Surface Area) of the two systems were checked. For the visualization of the MD trajectory files VMD v1.9.4 software (Humphrey et al. 1996) was used.

Toxicity and drug-likeliness analysis

We all know that in order to use a potential compound as a drug, the analysis of its toxicity and its drug-likeliness character is very essential. Protox-II (https://tox-new.charite.de/protox_II/) and SwissADME (http://www.swissadme.ch) are the two well-known servers for toxicity prediction and drug likeliness characterization of different type of molecules (Banerjee et al. 2018; Kumar et al. 2023). In order to do the drug likeliness property study, Lipinski’s rule of 5 (R_O5) study is very important. In this particular work, we carried out that study.

Essential dynamics or principal component analysis study

Apart from finding out different structural and conformational motions of an apo-protein and/or protein–ligand complex, essential dynamics study was also used to find out diagonalization and covariance matrices of different set of compounds by means of gmx_anaeig and gmx covar toolkits. In this particular piece of work apart from diagonalization and generation of covariance matrix, several eigenvectors and their eigenvalues were used to generate several collective motions respective to biomolecules, from which one can easily say that compound is stable or not.

Results and discussion

Analysis of docking result

By using the molecular docking method we can easily understand the substrate-molecule interactions which correlate well with several biological processes (El-Saghier et al. 2023; Abd El-Lateef et al. 2023). We performed molecular docking studies of 25 compounds with the two receptors (PDB ID 1O8A and PDB ID 3LV3). 1O8A-flunarizine complex (docking score = − 10.3 kcal/mol), and 3LV3-flunarizine complex (docking score = − 10.2 kcal/mol) were found to have the highest docking scores. Analysis of the structure reveals that human ACE inhibitor (PDB ID 1O8A) interacts with flunarizine via ASP 453, LYS 454, THR 282, TYR 523, PHE 457, GLN 281, GLN 369, GLN 162, CYS 370, PHE 527, VAL 379, GLU 376, TYR 520, VAL 380, ASP 377, ALA 354, THR 371 and THR 372 residues. The corresponding structures of protein–ligand complexes for both proteins have been represented in Figs. 2 and 3, respectively. From Fig. 2, it is clear that the flunarizine ligand interacts with the human ACE receptor (PDB ID 1O8A) via various interaction patterns e.g., van der Waals (vdW), π-anion, carbon-hydrogen, π-π, π-alkyl, etc. Among these interactions, hydrogen bonding interactions and vdW interactions played a crucial role concerning the stability of the ligand over the two receptors. Besides these, the respective PLIP (Protein–ligand interaction profiler) (Adasme et al. Jul. 2021) images are given in Fig. 4.

Fig. 2.

Different protein–ligand interactions of 1O8A-flunarizine complex

Fig. 3.

Different protein–ligand interactions of 3LV3-flunarizine complex

Fig. 4.

PLIP visualization of a 1O8A-flunarizine and b 3LV3-flunarizine complex

From the PLIP visualization, it is clear that human ACE inhibitor (PDB ID 1O8A) interacts with the ligand flunarizine via hydrophobic interactions e.g., THR282, ALA354, GLN369, VAL379, ASP453, PHE457, TYR523 and PHE527 and it forms a salt bridge with the residue GLU376. Whereas, 3LV3 interacts with the ligand flunarizine via several hydrophobic interactions e.g., THR190, LEU272, ARG273, TRP274, GLU275, hydrogen bond with VAL189 residue. The amino acid-hydrogen distance in the hydrogen bonding of the 3LV3-flunarizine complex remains within 3Ǻ, indicating the strong interacting hydrogen bonding residue. So, it can be demonstrated that a variety of weak interaction patterns stabilises the ligand within the targeted protein’s cavity. The relative distances of the bonds with the amino acid residues of the 1O8A-flunarizine and 3LV3-flunarizine complexes are given in Tables 2 and 3, respectively. Also, the corresponding interactions of all ligands with the two receptors have been given in Table S1 and Table S2 (supplementary information).

Table 2.

Different interacting residues and their distances (Ǻ) of the 1O8A-flunarizine complex

Residue	Amino acid	Distance (Ǻ)
Hydrophobic interactions
282A	THR	3.45
354A	ALA	3.86
369A	GLN	3.24
379A	VAL	3.50
453A	ASP	3.72
457A	PHE	3.30
523A	TYR	3.53
527A	PHE	3.42

Residue	Amino acid	Distance (Ǻ)
Salt bridges
376A	GLU	4.65

Table 3.

Different interacting residues and their distances (Ǻ) of 3LV3-flunarizne complex

Residue	Amino acid	Distance (Ǻ)
Hydrophobic interactions
190A	THR	3.66
272A	LEU	3.44
273A	ARG	3.87
274A	TRP	3.25
275A	GLU	3.98
275A	GLU	3.18

Residue	Amino acid	Distance H-A (Ǻ)	Distance D-A (Ǻ)
Hydrogen bonds
189A	VAL	2.07	2.95

Residue	Amino acid	Distance (Ǻ)
Salt bridges
275A	GLU	4.53
275A	GLU	5.38

Different experimental study showed ACE inhibitor as the potent anti-hypertensive agent in recent years. Ugwu and coworkers took three target proteins (PDB ID 2YDM, 2XDY and 3ZQZ) related to anti-hypertensive activity and docked MBPNPP with it. The authors also carried out MD simulation study of the docked complexes (Ugwu et al. 2023). Kalyan et al. did a structure-based analysis of enzyme-peptide interactions using molecular dynamics simulation, showing significant inhibitory properties of the peptides with the ACE inhibitor (Kalyan et al. 2021). Following these, we focused on flunarizine as the main potent inhibitor against another ACE receptor (PDB ID 1O8A) of this study showing the docking score of − 10.3 and − 10.2 respectively for 1O8A and 3LV3 receptors (see in Table 1). From Fig. 2, it is clear that 1O8A receptor interacts with the ligand flunarizine via different amino acid residues and due to higher no of hydrogen bonding and hydrophobic interactions, it shows variety of contacts indicating stable nature of interaction.

Prediction of ADMET

Whether a compound could be a potent drug or not, prediction of ADMET study is essential. From ADMET study, we can actually confirm the deviation of the compound from potent drug like behaviour (Singh et al. 2024). Thus, in order to find the drug-likeliness character of the 25 ligands prone to be drug, we had studied the ADMET (Adsorption, Distribution, Metabolism, Excretion, and Toxicity) properties of the ligands via Lipinski’s Rule of 5 characterization with the help of Swiss ADME server (Daina et al. 2017). We have found that all 25 ligands follow Lipinski’s rule of 5 i.e., molecular weight (MW) < 500 g/mol, number of H bond donors ≤ 5, number of H bond acceptors ≤ 10, and the logarithm of the partition coefficient (P) of a molecule between an aqueous and lipophilic phase (Log P) < 5, which confirm that these 25 ligands show the drug likeliness character. Furthermore, after the validation of various pharmacokinetic properties of the studied compounds, it suggests that the compounds were effectively absorbed by the gastrointestinal (GI) part with a low BBB (Blood Brain Barrier Permeability) value. The corresponding parameters related to the ADME properties of the compounds have been listed in Table S3. It is clear from Table S3 that except fosinopril and lercanidipine all other 23 compounds show higher value of GI absorption as a drug. Despite, lower GI absorption value, the compounds fosinopril and lercanidipine show Log P value < 5, which indicates that the two compounds are orally bioavailable and they have higher membrane permeability. Noteworthy, despite several limitations, the violation of Lipinski’s Ro5 is not uncommon, especially in larger molecules, which can cause potential issues in active transport. Overall, the favorable ADMET properties suggest promising pharmacokinetic profiles, with further areas of improvement.

Now, with the help of Protox-II web server, we have also analysed the toxicity study of the selected 25 compounds potent to be anti-hypertensive drugs. We found among 25 compounds, most of the compounds showed a low toxicity profile with a toxicity class of 4 or above, and some compounds showed a moderate toxicity profile with a toxicity class of 3. However, out of 25 compounds only 2 compounds, namely amlodipine and levamlodipine showed a toxicity class of 2 with a corresponding LD₅₀ (Lethal Dose 50) value of < 50 mg/kg, raising potential concerns for their clinical use as the drug. Despite showing a toxicity class of 2 and LD₅₀ value of < 50 mg/kg, the compounds amlodipine and levamlodipine showed no hepatotoxicity, carcinogenicity, immunotoxicity, or cytotoxicity profile. It is generally anticipated that a compound containing an electron-withdrawing group is often associated with mutagenicity risks e.g., lercanidipine. Despite these in-silico toxicity predictions, additional experimental validation via in-vitro and in-vivo studies will be critical to confirm these predictions. Detailed toxicity parameters of all 25 compounds are given in Table S4. Toxicity classes are distributed in 6 classes (level I-VI), among them toxicity classes of level I and II show higher toxicity than others. Thus by addressing the respective toxicity concerns of the 25 compounds, not only safety profiles of the compounds could be improved, but also it increase the clinical potential of the compounds to be used as drug.

Conceptual DFT study

As already stated, conceptual DFT is an important tool to analyse molecular and/or structural activity relationship (Geerlings et al. 2003; Kumar and Harbola 2024). Herein, initially the 25 ligands prone to be anti-hypertensive drugs were optimized using B3LYP (Becke, 3-parameter, Lee–Yang–Parr) functional (Kohn et al. 1996) by using the 6-31G basis set with the help of Gaussian 16 software (Frisch et al. 2016). We know the energy gap between frontier molecular orbitals of a molecule is estimated by E_LUMO − E_HOMO and the corresponding energy gap (ΔE_g) is directly related to the reactivity of the respective molecules. Among the 25 ligands, fosinopril showed the highest value of ΔE_g, and flunarizine and benidipine showed the lowest and corresponding 2nd lowest values, respectively, which correlate well with their docking results. The molecular dipole moment is another parameter from which we can also find out the chemical reactivity of the respective molecules. In this case, higher the dipole moment value, higher will be the chemical reactivity of the molecule. Nicardipine and cilnidipine were the two compounds for which the dipole moment values became > 10 Debye, which also correlates well with their respective molecular docking score which is > − 5.0 kcal/mol. Apart from these two molecular descriptive parameters, electronegativity is another parameter that denotes the efficacy of inhibition of the ligand with respect to the protein residue. Cilnidipine, nicardipine, nifedipine, barnidipine, benidipine and nimodipine were the compounds with the electronegativity value of greater than − 4.0 eV, which also correlate well with their docking results. Molecular docking score of these ligands were found to be > − 5.0 kcal/mol. The respective molecular descriptive parameters of the ligands prone to be anti-hypertensive drugs were tabulated in Table 4 and the structure and the charge distribution of all the ligands had been given in Table S5.

Table 4.

Conceptual DFT based molecular descriptors

Compound	Dipole moment (Debye)	EHOMO (eV)	ELUMO (eV)	ELUMO-EHOMO (eV)	Absolute hardness (η)	Global softness (S)	Electro negativity (χ)	Electrophilicity index (ω)
Alacepril	4.45	− 6.51	− 0.97	5.54	2.77	0.18	− 3.74	2.49
Amlodipine	7.56	− 5.50	− 1.45	4.05	2.03	0.25	− 3.48	2.98
Benazepril	3.61	− 5.92	− 0.60	5.32	2.66	0.19	− 3.26	1.99
Cilazapril	4.29	− 5.45	− 0.48	4.97	2.48	0.20	− 2.96	1.76
Cilnidipine	12.31	− 5.89	− 2.33	3.56	1.78	0.28	− 4.11	4.74
Clevidipine	6.06	− 5.95	− 1.83	4.12	2.06	0.24	− 3.88	3.65
Diltiazem	0.86	− 5.37	− 1.02	4.35	2.18	0.23	− 3.195	2.34
Enalapril	4.65	− 5.78	− 0.33	5.45	2.72	0.18	− 3.058	1.72
Flunarizine	2.71	− 5.65	− 0.88	4.77	2.38	0.21	− 3.264	2.24
Fosinopril	4.21	− 6.46	0.493	6.949	3.47	0.14	− 2.9815	1.28
Imidapril	7.66	− 5.83	− 0.84	4.99	2.49	0.20	− 3.335	4.46
Lercanidipine	6.70	− 5.36	− 2.69	2.67	1.34	0.37	− 4.025	6.06
Levamlodipine	7.45	− 5.50	− 1.45	4.05	2.03	0.25	− 3.478	2.99
Moexipril	5.72	− 5.76	− 0.46	5.30	2.65	0.18	− 3.108	1.82
Nicardipine	13.68	− 5.65	− 3.29	2.36	1.18	0.42	− 4.466	8.46
Nifedipine	3.34	− 5.65	− 2.41	3.24	1.62	0.31	− 4.0305	5.00
Nisoldipine	3.51	− 5.65	− 2.39	3.26	1.63	0.30	− 4.019	4.96
Perindopril	2.69	− 5.72	− 0.82	4.90	2.45	0.20	− 3.267	2.18
Barnidipine	5.17	− 5.68	− 2.67	3.01	1.50	0.33	− 4.1735	5.78
Benidipine	6.26	− 5.78	− 2.55	3.23	1.62	0.31	− 4.1685	5.38
Quinapril	4.03	− 5.94	− 0.47	5.47	2.73	0.18	− 3.2065	1.88
Ramipril	5.28	− 6.12	− 0.40	5.72	2.86	0.17	− 3.262	1.86
Verapamil	5.76	− 5.20	− 0.04	5.16	2.58	0.19	− 2.6215	1.33
Zofenopril	6.25	− 6.39	− 1.53	4.857	2.43	0.20	− 3.9615	3.23
Nimodipine	7.56	− 5.86	− 2.52	3.34	1.67	0.29	− 4.1875	5.25

MESP (molecular electrostatic potential) analysis

Molecular electrostatic potential (MESP) map identifies the negative, positive and neutral electrostatic potential zones in a molecule. Herein, the blue region of the MESP map denotes the electropositive region (electron deficient region) of a particular compound and whereas the red region denotes the particles with a positive electrostatic potential region of the compound i.e., highly electronegative region of the compound. By determining the electrophilic and nucleophilic regions of the compounds, we can easily find the regions with different electron density in a compound, which actually helps us to assess a detailed account of interactions of the compounds with the receptors (Rizwana and Nigar 2020). From Fig. 5, it is clear that the energy of the flunarizine ligand is given in the range ± 3.593e⁻², whereas for the benidipine ligand, the given range is ± 6.993e⁻². Flunarizine and benidipine are the two highest-scored ligands among the 25 ligands.

Fig. 5.

Molecular electrostatic potential (MESP) maps of a flunarizine and b benidipine

Analysis of MD simulation results of human ACE receptor complex (1O8A-flunarizine)

Molecular dynamics (MD) simulations are used to understand the stability of several proteins, molecular recognition and several conformational changes related to protein–ligand interactions. From the previous docking study, we have seen that flunarizine mainly interacts with the receptor 1O8A via ASP 453, LYS 454, THR 282, TYR 523, PHE 457, GLN 281, GLN 369, GLN 162, CYS 370, PHE 527, VAL 379, GLU 376, TYR 520, VAL 380, ASP 377, ALA 354, THR 371 and THR 372 residues. Evaluation of root mean square deviation (RMSD) of flunarizine with human ACE receptor (PDB ID 1O8A) is shown in Fig. 6 which clearly reveals that despite initial fluctuations of the 1O8A-flunarizine complex till 40 ns of MD simulation study, it becomes stable afterward and remains stable till 100 ns, which actually indicates the stable behavior of the protein–ligand complex and the RMSD value remains within 1.5Ǻ. On the other hand, from the corresponding root mean square fluctuation (RMSF) plot as depicted in Fig. 7, it is confirmed that the fluctuations of residues in the docked structure are more compared to the undocked one, however, the fluctuations are less (for both free protein and for protein–ligand complexes, the fluctuations remain with 0.25 ± 0.5 nm) and indicating the stability of the protein–ligand complex. We can thus say that the root mean square fluctuation (RMSF) plot correlates well with the root mean square deviation (RMSD) plot.

Fig. 6.

Root mean square deviation (RMSD) plot of free 1O8A and 1O8A-flunarizine complex

Fig. 7.

Root mean square fluctuation (RMSF) plot of free 1O8A and 1O8A-flunarizine complex

Besides Figs. 6 and 7, we have done the radius of gyration (R_g) plot of free human ACE receptor (PDB ID 1O8A) and the receptor-ligand complex (1O8A-flunarizine) in Fig. 8, demonstrating how the radius of gyration changes for free receptor and the docked complex. We know that the radius of gyration measures the overall compactness of the system, hence from Fig. 8 we can conclude that the docked complex is less compact than the undocked one, however, the protein–ligand complex (1O8A-flunarizine) shows an acceptable value of R_g < 2.5 nm, which again indicates the stability of the protein–ligand complex. All the previous observations i.e., both the root mean square deviation (RMSD) and root mean square fluctuation (RMSF) correlate with the radius of gyration (R_g) result. Like the radius of gyration (R_g), Solvent Accessible Surface Area (SASA) also deals with the compactness of the system within the solvent system. Here, Fig. 9 represents the SASA plot of the two systems, i.e., free human ACE receptor (PDB ID 1O8A) and the docked complexes of 1O8A and flunarizine. From Fig. 9 it is clear that, the docked complex of 1O8A-flunarizine complex fluctuates more than the undocked one, indicating the less compactness of the docked structure, however both the structures maintain a stable shape within 100 ns time frame. Thus, from this analysis we can say that all the plots correlate well with one another, indicating some structural and conformational change nature of the docked structure. However, the complex remains stable within the 100 ns of time frame. The change of binding modes of the 1O8A-flunarizine docked complex before and after MD is shown later.

Fig. 8.

Radius of gyration plot of free 1O8A and 1O8A-flunarizine complex

Fig. 9.

Solvent accessible surface area (SASA) plot of free 1O8A and 1O8A-flunarizine complex

Analysis of MD simulation results of HLA complex human voltage dependent receptor complex (3LV3-flunarizine)

From the molecular docking study of the 3LV3-flunarizine complex, we demonstrated that 3LV3 mainly reacts with the ligand flunarizine via THR190, TRP274, ARG273, HIS188, GLU275, LEU272, THR271, and VAL189 residues. Now, to measure the stability of the 3LV3-flunarizine complex, we have again plotted several parameters like RMSD, RMSF, R_OG, and SASA related to that complex. The RMSD plot of the 3LV3-flunarizine complex is shown in Fig. 10. From the figure, it is clear that, at the beginning of the molecular dynamics (MD) simulation process, the docked 3LV3-flunarizine complex shows higher fluctuation compared to the undocked free 3LV3 protein, indicating higher flexibility of the protein–ligand complex structure. However, the complex remains stable throughout the 100 ns of the molecular dynamics (MD) simulation process and the RMSD value remains within < 2Ǻ. Besides, the corresponding RMSF plot is shown in Fig. 11, demonstrating that the 3LV3-flunarizine complex shows higher fluctuation compared to the undocked free protein, resembling the RMSD plot, however, fluctuations of both the free protein and the protein–ligand complex are less and remain within 0–0.5 nm, indicating the stability of the complex. Both the figures ascertain the stability of the protein–ligand complex (3LV3-flunarizine).

Fig. 10.

Root mean square deviation (RMSD) plot of free 3LV3 and 3LV3-flunarizine complex

Fig. 11.

Root mean square fluctuation (RMSF) plot of free 3LV3 and 3LV3-flunarizine complex

Furthermore, we have also shown the R_g and SASA plots of the free 3LV3 receptor and 3LV3-flunarizine complex in Figs. 12 and 13, respectively. The R_g plots of free 3LV3 receptor and 3LV3-flunarizine complex show similar characteristics to the RMSD plot. From Fig. 12, it can be depicted that the 3LV3-flunarizine complex shows very small fluctuation than the free 3LV3 receptor, indicating some less compactness of the protein–ligand complex, however, that remains within the range of ± 2.5 nm till 100 ns of MD simulation process, which again indicates the stability of the protein–ligand complex. In order to do further validation of the results, we have made SASA plots of both free 3LV3 receptor and 3LV3-flunarizine complex, which also correlate well with their previous results. The binding modes of the 3LV3-flunarizine complex before and after MD is depicted in the next section.

Fig. 12.

Radius of gyration (R_g) plot of free 3LV3 and 3LV3-flunarizine complex

Fig. 13.

Solvent accessible surface area (SASA) plot of free 3LV3 and 3LV3-flunarizine complex

Potamitis and co-workers examined the docking and MD simulation study of the ligand valsartan (similar type of ligand like flunarizine) with the AT₁ receptor in both explicit and implicit solvent medium, in order to find out the character of the ligand valsartan over the AT₁ receptor. They had shown that valsartan formed several H bonding and hydrophobic interactions with the receptor, indicating the stable nature of the complex (Potamitis et al. 2009). We all know that in order to take care of hypertension (HTN) measurement, the 1st line treatment along with the thiazide dieuretics and calcium channel blockers, ACE is an important agent. Kumar et al. took ACE-captopril complex (PDB ID 1UZF) as the receptor and docked ADMQ ligand with it. Finally with the help of MD simulation study, they had proven the effect of the ligand ADMQ over the ACE-captopril receptor (Kumar et al. 2015). Based on this observation we took another ACE inhibitor (PDB ID 1O8A) and L type calcium channel and docked potent 25 ligands with them and after doing DFT and MD simulation study we found flunarizine as the best potent inhibitor among them. Saddala et al. did pharmacophore based computational modelling method in order to screen different compounds, which can act as novel CCB (calcium channel blockers). Despite of computational approach, they also did in-vitro and in-vivo studies respectively, in order to find out the inhibitory activity of the ligand over the receptor (Saddala et al. 2020). Based on this fact, we took another L type CCB (PDB ID 3LV3) and docked 25 ligands with the receptor and again after all the possible computational exposures we got flunarizine again as the best potent inhibitor.

Structural and conformational changes of the complexes before and after MD simulation

Here, we discuss the conformational changes of the studied protein–ligand complexes before and after MD simulations. From Fig. 14, it is absolutely clear that the 1O8A-flunarizine complex remains stable before and after the MD simulation process. But due to some structural and conformational changes, the interaction residues changes, which correlates well with the previously said RMSD, RMSF, R_g and SASA plots. However, due to remaining of hydrogen bonding in 1O8A-flunarizine complex before and after MD, it shows a stable nature, and the ligand also remains within the cavity of the receptor molecule.

Fig. 14.

Change of conformations of 1O8A-flunarizine complex a before MD b after MD

Figure 15 reveals that despite of some conformational and structural changes, 3LV3-flunarizine complex remains stable like former 1O8A-flunarizine complex both before and after the MD simulation process. Thus, from these both structural and graphical observations, it is clear that flunarizine is a better potent inhibitor among these 25 ligands and in-vitro and in-vivo observations can further confirm it.

Fig. 15.

Change of conformation of 3LV3-flunarizine a before MD b after MD

Principal component analysis (PCA) of 1O8A-flunarizine and 3LV3-flunarizine complexes

In order to describe protein dynamics of a particular apo-protein or protein–ligand complex, principal component analysis (PCA) is used. With the help of the PCA, one can easily identify conformational space and transition in the apo-protein and the docked complex during the 100 ns of the MD simulation process (Rana et al. 2023). The corresponding figures of PCA are given in Fig. 16. The major contribution to the protein–ligand fluctuations is shown by the first three modes of eigenvalues. From Fig. 16, it is clear that the first three principal components (PCs) account for the most significant conformational changes in the free protein and the protein–ligand complexes (free 1O8A, free 3LV3, 1O8A-flunarizine and 3LV3-flunarizine). After the projection of the MD trajectory files onto PC1 to PC3, we found stable distinct clusters, indicating that the compound maintained a stable conformation with limited transition between states. With the help of these PCA plots, we verified the results of the RMSD, RMSF, R_g, and SASA plots of both the free protein and the protein–ligand complexes, and we found a significant correlation between the MD simulation plots and the PCA result. In principal component analysis (PCA), PC1 (principal component 1) plays an important role by measuring large scale motions of the protein binding pockets of the protein–ligand complex. With the help of PC2 (principal component 2) and PC3 (principal component 3) small scale adjustment near the active site of the protein was captured. From Fig. 16, we can observe stable clustering of the complex in PCA space, which confirms the hypothesis of tightly bound and stable dynamic interaction. Thus, from these PCA studies, we can get valuable insights into the role of dynamic structural changes in ligand efficacy.

Fig. 16.

Principal component analysis (PCA) plots of a free 1O8A and 1O8A-flunarizine complex b free 3LV3 and 3LV3-flunarizine complex

Structural trajectory analysis of the MD simulation study of 1O8A-flunarizine and 3LV3-flunarizine complexes

It is well established that, in order to validate a docking protocol by means of the changes in structural and conformational dynamics of the docked complex, structural trajectory analysis of MD simulation of the docked complexes is very essential. By means of experimental techniques, one cannot identify the structural and conformational changes within a macromolecular system (Mohd et al. 2019; Dahiya et al. 2019). For analysing the conformational changes of the protein–ligand structure, we have taken snapshots of the trajectories of the 1O8A-flunarizine complex at every 10 ns of time frame. As evidenced by the RMSD plot (Fig. 6), the convergence time of the 1O8A-flunarizine system is 50 ns, thus, we take the snapshots from 50 ns time frame to 100 ns of time frame. Figure 17 denotes the snapshots of the 1O8A-flunarizine complex which indicate that the ligand remains within the binding pocket from 50 to 100 ns time, demonstrating its stable behaviour.

Fig. 17.

Snapshots of trajectories of 1O8A-flunarizine complex in every 10 ns interval

On the other hand, recalling Fig. 10, the RMSD plot of 3lv3 free and 3LV3-flunarizine complex, it is evident that the protein–ligand complex shows a stable behaviour throughout the 100 ns of MD simulation time. In order to further confirm the stability throughout the 100 ns of MD simulation process, the structural trajectories of the 3LV3-flunarizine complex were taken at an interval of 20 ns. The structural trajectories (Fig. 18) confirm that the ligand remains within the binding pocket of the protein throughout the 100 ns of MD simulation process.

Fig. 18.

Snapshots of trajectories of the 3LV3-flunarizine complex in every 20 ns interval

Conclusion

Hypertension and cardiovascular diseases (CVD_s) belong to two main classes of diseases that have caused global healthcare problems in recent years. These two diseases are also associated with high blood pressure disease. Among the risk factors related to the development of cardiovascular diseases, hypertension is one of them (Lastname et al. 2008). However, hypertension and cardiovascular diseases can be prevented by means of ACE and renin angiotensin converting (RAC) enzyme treatment. In this work, we had taken two receptors corresponding to the anti-hypertensive activity (PDB ID 1O8A and 3LV3) and docked 25 ligands with each of the receptors. From molecular docking structural visualization, we found various important interacting residues e.g., hydrogen bonding, hydrophobic interaction, van der Waals interaction, etc., of the ligand flunarizine concerning 1O8A and 3LV3 receptors that indicate the stable nature of binding of the ligand flunarizine concerning the two receptors. We optimized all the biologically active ligands that pretended to be anti-hypertensive drugs by means of 6-31G basis set and B3LYP functional, from which we can easily justify that different global reactivity parameters e.g., E_HOMO, E_LUMO, chemical potential, electronegativity (χ), etc. directly correlated with their respective docking scores. Finally, we have done molecular dynamics (MD) simulation study of the two best docked complexes i.e., 1O8A-flunarizine and 3LV3-flunarizine complexes. We have plotted RMSD, RMSF, R_g and SASA of the two complexes. After analysis of conformational dynamics, we have found a significant impact of the ligand flunarizine over the two receptors, indicating its high drug-like character. However, further in-vitro and in-vivo studies are required for further assessment.

Supplementary Information

Below is the link to the electronic supplementary material.

Authors contributions

S.G. performed the computations, prepared the initial draft. A.P. revised the manuscript to final form. All authors reviewed the manuscript.

Funding

This work was funded DST-FIST, Delhi, Govt. of India vide project sanction letter no. SR/FST/CS-I/2022/247 and UGC-BSR Research Start-Up-Grant (No.F.30-557/2021 (BSR)).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Conflict of interest

The authors declare no competing interests.

Ethics and consent to participate

Not applicable.