Computational Analysis of Phenylpropyl Amine Derivatives as Potential Calcium Channel Blockers Using an Ancestral Bacterial Channel as a Structural Proxy

Phenylpropylamines act as calcium channel blockers and are used to treat cardiovascular diseases. Verapamil is a typical example. They block calcium influx into cardiac smooth muscle cells, thereby reducing heart rate. They also have vasodilatory effects. Amlodipine is a calcium channel blocker that acts as a vasodilator by blocking calcium channels. It is a complex drug molecule containing a propoxy side chain. The propyl amino group contributes to a molecule's lipophilicity, thereby enhancing its penetration of the blood-brain barrier^1-3. The amino (-NH2) group or the oxo group can act as a hydrogen-bond donor or acceptor in interactions with biological targets. The propyl amino group can fit into the hydrophobic pockets of several biological receptors, enhancing biological activity. Calcium channel blockers such as amlodipine and verapamil are frequently used to treat angina pectoris, hypertension, and certain cardiac arrhythmias.^4-5 Derivatives of phenyl propylamines are a significant family of bioactive substances with a wide range of pharmacological actions. Both aromatic and amino functional groups in the phenyl propylamine scaffold promote hydrophobic, hydrogen bonding, and electrostatic interactions with biological targets. These molecules can adopt favorable conformations within receptor binding pockets due to the flexibility of the propylamine chain, thereby improving ligand–receptor interactions. Because structural changes to the phenyl propylamine nucleus have been shown to affect pharmacokinetic and pharmacodynamic characteristics, this scaffold is a desirable option for the creation of new medicinal compounds. The crystal structure of the CavAb–Amlodipine complex (PDB ID: 5KMD) provides important insights into the molecular underpinnings of calcium channel blockade. A helpful structural model for understanding ligand interactions with calcium channels is CavAb, a voltage-gated calcium channel from bacteria. The co-crystallized amlodipine molecule makes it easier to evaluate possible calcium channel blockers using computational methods by identifying significant binding sites within the channel pore. Identifying substances with favorable binding properties and potential antihypertensive efficacy can be facilitated by analyzing ligand interactions with the 5KMD receptor. In contemporary pharmaceutical research, computational drug discovery methods such as molecular docking and molecular dynamics simulations have become essential tools. Molecular docking helps identify prospective lead compounds by estimating binding affinity and predicting a ligand's preferred binding orientation inside a target receptor.^6-7 By assessing the stability and dynamic behavior of protein–ligand complexes under physiological settings, molecular dynamics simulations supplement docking investigations. When combined, these methods reduce the time and expense of experimental screening while providing useful information on molecular recognition, binding stability, and structure–activity relationships. These propylamine chemicals can be tested in silico for their β-adrenergic binding properties, BBB permeability, and hydrophilic interactions⁸. The 5-KMD receptor is a transporter protein (CavAb in Complex with Amlodipine). Phenylalkylamines and benzothiazepines are primarily used for the treatment of cardiac arrhythmias, as they physically block the pores of the calcium channel, which is inhibited by dihydropyridines and phenylalkylamines⁹.

Molecular dynamics

The GROMACS molecular dynamics online web server enables researchers to run simulations. Molecular docking and dynamic simulations of different ligand-target interactions using GROMACS are widely used open-source Software in chemistry ¹⁰. Several advanced techniques are available for regulating temperature/pressure stimulation, employing various geometric constraints and utilizing both explicit and implicit solvents.

Molecular docking

GNINA is an open-source molecular docking tool that integrates several convolutional neural networks to enhance the accuracy of protein-ligand binding predictions. The GNINA online web server can be accessed with a NeuroSnap account. It integrates Convolutional Neural Networks (CNNs) to improve performance. It supports flexible docking and assesses machine learning11. Calcium channel blockers are widely used to control blood pressure and manage angina symptoms. They are well-tolerated and safe medications12. Verapamil is an example of a phenylpropanolamine compound. These compounds are also used to treat coronary heart spasm.

MATERIALS AND METHODS

Docking was performed using GNINA. Molecular dynamics were performed using GROMACS. ADME was evaluated using Swiss ADME. The target receptor, RCSBPDB:5KMD, was downloaded from the RCSB Protein Data Bank (PDB) web server (Figure:1-2). The ligands were downloaded from the PubChem database in 3D SDF format. The laptop, equipped with an Intel i7 processor and Windows 11 OS, featured Iris Graphics. The molecular dynamics simulations were performed using GROMACS and the Neurosnap web server, which uses the GROMACS framework to simulate solvent-solute interactions13. The Neurosnap web server enables you to run GROMACS M.D. with an account registered with the software company. It simulates a broad range of molecular systems, providing RMSD, RMSF, and radius of gyration (Figure:3).

GNINA

It is an open-source molecular docking tool that integrates Convolutional Neural Networks (CNNs) to enhance the accuracy of protein-ligand binding predictions¹⁴. The Receptor PPDB form was uploaded and optimized. Ligands were uploaded over time, and docking was performed. The optimized results were obtained as a CSV file. The CNN Pose score, CNN affinity (kcal/mol), and intramolecular energy were provided at ¹⁵.

ADME Study

Swiss-ADME was used to obtain log P, BBB penetration, and synthetic accessibility. It is an online web server where ligand structures can be uploaded to obtain the results. The chemical structure was uploaded as a SMILES string. The results were obtained after the run time. The The bioavailability radar and boiled-egg diagram summarized the results. The values obtained are shown in Table ¹⁶.

Protox for Toxicity Prediction

The Protox web server was utilized for toxicity prediction. The ligand of interest was tested by uploading its chemical structure; the output showed variation in toxicity parameters¹⁷.

Dynamic Bind (Molecular dynamics (MD) simulations)

In contrast to conventional docking techniques that consider proteins as static and computationally costly molecular dynamics (MD) simulations, Dynamic-Bind offers a "dynamic docking" method for examining protein–ligand interactions. Its ability to study large-scale protein dynamics has important implications for drug discovery, particularly for compounds targeting cryptic pockets ¹⁸.

RESULTS

Table 1. Data from molecular docking of different phenylpropylamines

* Convolutional Neural Network (CNN) that evaluates the quality of a given protein-ligand pose leverages the power of deep learning to evaluate protein-ligand poses more accurately

** CNN affinity score attempts to quantify the strength of the interaction, often as a binding energy (e.g., in kcal/mol)

*** The more negative the value, the more the biological activity

*** The less the intramolecular energy, the more stable the binding and the better the results

Table. 2. Results of the Molecular dynamics simulation study of the 5KMD receptor

*Mean pLDDT (predicted Local Distance Difference Test), a higher score indicates greater confidence in the predicted local structure ** The "uniqueness" of each conformation sampled during the simulation contributes to a more comprehensive understanding of the entire binding

*** RMSD (Root Mean Square Deviation) is a critical metric used to quantify the average distance between the atoms. A lower RMSD value indicates a higher degree of similarity between the structures.

Table. 3 ADME Parameters of selected ligand