DFT, Docking, and Molecular Dynamics Investigation of Amazonian Alkaloids on the Dopamine D(2) Receptor Protein

Abstract

Alkaloids are nitrogen-containing compounds with diverse pharmacological activities, many of which remain underexplored, particularly those from Amazonian biodiversity. This study investigates the interaction of selected Annonaceae alkaloids with the dopamine D(2) receptor protein data bank (PDB ID: 6CM4), a key target in neurological and psychiatric disorders. An integrated computational approach combining absorption, distribution, metabolism, excretion, and toxicity (ADMET) prediction, molecular docking, molecular dynamics (MD) simulations, molecular mechanics/Poisson-Boltzmann surface area (MM/PBSA) binding free energy calculations, and density functional theory (DFT) analysis was employed to evaluate their pharmacological potential. Molecular dynamics simulations (100 ns) demonstrated that pallidine, orientaline, and N-methylcoclaurine form stable complexes, characterized by reduced conformational fluctuations, increased structural compactness, and persistent hydrogen bonding. Principal component analysis and free energy landscape indicated that these ligands restrict receptor conformational space, stabilizing low-energy states. MM/PBSA calculations showed that binding is predominantly driven by van der Waals interactions, while polar solvation opposes complex formation. Density functional theory revealed that compounds with lower highest occupied molecular orbital and lowest unoccupied molecular orbital (HOMO-LUMO) gaps, particularly pallidine (3.92 eV), exhibit enhanced electronic adaptability, favoring ligand-receptor interactions. Overall, pallidine, orientaline, and N-methylcoclaurine emerge as promising lead compounds for dopamine D(2) receptor modulation. These findings provide mechanistic insights into ligand-receptor recognition and establish a rational basis for future experimental validation and structure-activity relationship optimization of Amazonian alkaloids as potential neuropharmacological agents.

Keywords:
dopamine D2 receptor; Amazonian alkaloids; molecular dynamics simulation; MM/PBSA binding free energy; dopaminergic activity; GPCR-ligand interactions

Introduction

Alkaloids are a diverse class of nitrogenous organic compounds, primarily derived from plant secondary metabolism but also found in fungi, bacteria, and animals.¹,² Composed mainly of carbon, hydrogen, nitrogen, and oxygen, they exhibit broad pharmacological activities, including anticancer, antimicrobial, antidepressant, antimalarial, anticholinergic, anesthetic, and vasodilator effects.³-⁶ Over 6,000 alkaloids have been identified from approximately 4,000 plant species, with many, particularly those from the region Amazon biodiversity, remaining underexplored in terms of chemical and biological properties.⁶

This untapped potential positions alkaloids as a promising source for novel drug development. The compounds investigated here, isolated from Annonaceae species by GEQBiom research group (Grupo de Estudo de Química de Biomoléculas - Biomolecular Chemistry Research Group) and selected based on therapeutic potential rankings (as detailed in the methodology), include pallidine (morphinanedienone subclass); N-methylcoclaurine, juziphine, orientaline, reticuline, magnococline, and coclaurine (benzyltetrahydroisoquinoline subclass); and 13-hydroxy-discretinine, discretamine, 13-hydroxy-2,3,9,10-tetramethoxyprotoberberine, coreximine, isocoreximine, stepholidine, discretine, corytenchine, xylopinine, corypalmine, and 7,8-dihydroxypalmatine (tetrahydroprotoberberine subclass) (Figure S1, Supplementary Information (SI) section). The Annonaceae family, comprising about 112 genera and 2,440 tropical/subtropical species, has long been valued for medicinal and edible uses.⁷-⁸ Phytochemical analyses reveal diverse secondary metabolites, including monoterpenes, diterpenes, triterpenes, lignans, flavonoids, and especially isoquinoline-derived alkaloids.⁹-¹³

Alkaloids from Annonaceae have shown notable biological activities, such as anti-inflammatory and urease inhibitory,¹⁴,¹⁵ trypanocidal,¹⁶,¹⁷ leishmanicidal,¹⁷,¹⁸ antimalarial,¹⁰,¹⁹ antimicrobial,¹⁰,²⁰ antioxidant and antirheumatic,¹⁵,²¹ and cytotoxic effects against tumor cell lines.¹⁰,¹²,¹⁷,²² This pharmacological versatility underscores their therapeutic promise and the need to explore new targets.

In this context, based on the therapeutic relevance of the dopamine D(2) receptor and the chemical diversity of Amazonian alkaloids, this study aims to systematically investigate the interaction mechanisms between selected Annonaceae-derived alkaloids and the D(2) receptor protein data bank (PDB ID: 6CM4) by an integrated computational approach. The objective is to elucidate the structural, dynamic, and electronic determinants governing ligand-receptor recognition, identify potential lead compounds, and provide mechanistic insights to support future experimental validation and structure-activity relationship optimization for dopaminergic modulation. Despite the widespread use of combined molecular docking, molecular dynamics, and density functional theory (DFT) approaches in drug discovery,²³,²⁴ the novelty of the present study lies in the integrated and system-specific investigation of Amazonian alkaloids targeting the dopamine D(2) receptor, a combination that remains largely unexplored. Unlike previous studies that typically focus on isolated computational methods or well-characterized ligand classes, this work provides a multi-level correlation between pharmacokinetic properties absorption, distribution, metabolism, excretion, and toxicity (ADMET), molecular dynamics (MD), binding energetics molecular mechanics/Poisson-Boltzmann surface area (MM/PBSA), and electronic structure (DFT) within a unified framework. Importantly, the study goes beyond descriptive analysis by establishing a mechanistic link between electronic descriptors (e.g., highest occupied molecular orbital and lowest unoccupied molecular orbital (HOMO-LUMO) gap, electrophilicity) and dynamic binding behavior, demonstrating how electronic adaptability influences interaction persistence and receptor stabilization. In addition, the identification of Amazonian alkaloids as potential modulators of the D(2) receptor expands the chemical space of dopaminergic ligands, which is still predominantly dominated by synthetic scaffolds. This approach correlates electronic donor-acceptor capabilities with stabilization of orthosteric pocket interactions (e.g., with Asp114 and Ser193), elucidating dynamic, energetic, and structural properties of alkaloid-receptor complexes to identify therapeutic leads and advance structure-activity relationships (SAR). Therefore, the originality of this work resides not only in the computational workflow itself, but in the integration of electronic, structural, and dynamic insights applied to a biologically and chemically underexplored class of natural compounds, providing a rational basis for future structure-activity relationship studies and experimental validation.

Experimental

Toxicity analysis and bioactivity prediction

To explore the interest in the discovery of bioactive drugs, we calculated physicochemical descriptors, predicted ADMET parameters, pharmacokinetic properties, drug-likeness, and the affinity of the molecule as a medicinal chemistry agent for the alkaloids in question using the available tools on SwissADME.²⁵-²⁷ The analysis of ADMET, medicinal chemistry, and drug- and lead-like properties was conducted using online tools such as SwissADME,²⁷ pkCSM,²⁸ ProTox-II²⁹ and OSIRIS property explorer.³⁰ These tools were employed to evaluate the potential toxicity of drugs and compounds designed for human use. Additionally, a bioavailability radar analysis was performed to verify the similarity between the drug and the bioavailability of the identified compound (Figures S2 and S3, SI section).

Target protein selection

Following ADMET and drug-likeness predictions, the selected alkaloids satisfied key criteria for bioactive drug candidates. Potential therapeutic targets were identified using ligand-based approaches that leverage chemical similarity to known active compounds. The similarity ensemble approach (SEA)³¹ and SwissTargetPrediction²⁷ platforms were employed for this purpose.

These tools predict targets by comparing query molecules to libraries of compounds with established bioactivities against various proteins. SEA relates targets through set-wise ligand similarity, enabling rapid screening and cross-target mapping. SwissTargetPrediction combines 2D and 3D similarity searches against extensive databases of known actives (over 376,000 compounds on > 3,000 targets in humans, mice, and rats) to rank probable macromolecular targets.

This ligand-based strategy is widely validated in drug discovery for virtual screening, repurposing known compounds, and prioritizing targets based on structural analogies to reference ligands.³²-³⁴ Comparative analysis via SEA and SwissTargetPrediction identified the dopamine D(2) receptor (PDB ID: 6CM4) as the top-ranked target for the highest-scoring alkaloids.

Dopamine, a key central nervous system neurotransmitter, modulates diverse physiological and behavioral processes through its receptors, including the D(2) subtype. Modulating D(2) receptor activity (via agonism or antagonism) holds therapeutic relevance for neuropsychiatric and neurological disorders, such as: (i) schizophrenia: D(2) receptor blockade alleviates positive symptoms (e.g., hallucinations, delusions), forming the basis of most antipsychotic therapies,`³⁵-³⁸ (ii) attention deficit hyperactivity disorder (ADHD): D(2) receptor activation may improve attention and impulsivity control,³⁹,⁴⁰ (iii) cognitive enhancement: selective D(2) modulation supports working memory and learning, with potential in neurodegenerative or cognitive disorders,⁴¹,⁴² (iv) mood disorders (e.g., depression, bipolar disorder): dopaminergic pathways regulate mood, and D(2) modulation offers benefits in mood stabilization,⁴³,⁴⁴ (v) movement disorders (e.g., Parkinson’s disease): although dopamine deficiency predominates, selective D(2) modulators (often agonists) aid symptom management.⁴⁵ This target selection guided subsequent structure-based investigations of alkaloid-D(2) receptor interactions.

Calculation and interaction of docking

In this study, the receptor was treated as rigid during docking, a widely adopted approximation that provides reliable identification of binding modes and relative interaction patterns at a reduced computational cost. This approach is particularly appropriate given the structurally well-defined binding site of the target protein.⁴⁶ Additionally, potential limitations associated with receptor rigidity were mitigated by subsequent molecular dynamics simulations, which allow conformational relaxation and dynamic refinement of ligand-protein interactions.

The dopaminergic D(2) receptor structure (PDB ID: 6CM4)⁴⁷ was retrieved from the protein data bank⁴⁸ and prepared using Schroedinger Maestro 2023-2.⁴⁹ Preparation involved removing water and co-factors, adding hydrogens at pH 7.4, assigning partial charges, and modeling missing residues. Simultaneously, alkaloid ligands were processed in AutoDock Tools 1.5.7 (The Scripps Research Institute, La Jolla, CA, USA, 2009) to assign Gasteiger charges and define rotatable bonds, converting them to .pdbqt format.

The search space was centered on the co-crystallized risperidone site using AutoGrid FR 2019 v1.0 (The Scripps Research Institute, La Jolla, CA, USA) with coordinates x = 9.177, y = 5.492, z = −8.665 and dimensions of 27.0 × 30.0 × 35.25 Å. Molecular docking was executed via AutoDock Vina 1.2.5⁵⁰ with an exhaustiveness of 96, maintaining a rigid receptor and flexible ligands. The best-scoring conformations (ΔG_binding) were converted using OpenBabel 2019,⁵¹ and complexes were analyzed in Maestro and BIOVIA discovery studio 2021 developed by Dassault Systèmes BIOVIA. For enhanced affinity prediction, the complexes were rescored using the KDeep AI-based platform.⁵²

Protocol validation was performed by redocking the native ligand (risperidone). Structural alignment via LS-Align⁵³ yielded an RMSD of 1.5743 Å, satisfying the literature-recommended threshold of < 2 Å.⁵⁴ This result (Figure 1) confirms the reliability of the docking parameters for investigating the target alkaloids.

Figure 1.
Redocking result for the 6CM4 receptor and the risperidone ligand, where the RMSD was 1.5743Å.

Ranking and identification of lead compounds

A highly significant stage in drug development involved the ranking and identification/optimization of lead compounds. Various calculations, encompassing docking scores, interaction analysis, and ADMET analysis, including molecular weight (MW), hydrogen bond donors (HBD), hydrogen bond acceptors (HBA), octanol-water partition coefficient (logP), polar surface area (PSA), rotatable bonds, and rings, were conducted to identify lead compounds that meet the criteria established by Ghose filter, Veber and Egan rules, and the rule of five (ROF).⁵⁵

Compounds demonstrating good interaction, the best fitting score, and binding affinity were selected as potential actuators (agonists or antagonists) for the target protein. Additionally, the remaining ranked compounds are subject to further analysis, considering their specific advantages highlighted in the scores, thus representing secondary options for drug development.

DFT calculations

To investigate the reactivity and some other characteristics of alkaloids through the HOMO-LUMO orbitals analysis, DFT calculations were performed using Gaussview 5.0 and Gaussian 09 Program (Revision E.01),⁵⁶,⁵⁷ employing the DFT approach and the 6-31G basis set, along with the Becke-3-Lee-Yang-Parr (B3LYP) functional.⁵⁸ It is important to emphasize that the DFT calculations in this work were not intended to provide highly accurate thermochemical or spectroscopic data, but rather to serve as a supporting and qualitative role within the broader framework of the MD investigation. In particular, DFT was employed to (i) obtain reliable equilibrium geometries and (ii) provide qualitative insight into frontier orbital distributions, supporting the interpretation of MD-derived structural and interaction patterns. While larger basis sets including polarization functions, such as 6-311(d,p), are known to improve the quantitative description of electronic properties, especially for heteroatom-containing systems, the choice of the 6-31G basis set reflects a deliberate balance between computational efficiency and chemical reliability, particularly for the relatively large alkaloid systems considered here. The B3LYP functional combined with split-valence basis sets has been extensively validated and widely applied in studies of organic and bioactive molecules, providing robust geometries and qualitatively consistent electronic structures.⁵⁸ Furthermore, it should be noted that the molecular dynamics simulations were conducted using classical force fields, whose parametrization does not directly depend on high-level quantum mechanical descriptions. Therefore, increasing the level of theory in the DFT calculations would not translate into a meaningful improvement in the MD results, which constitute the central objective of this study. The DFT-optimized structures have been added to the Supplementary Information (Figures S4 and S5, SI section).

Molecular dynamics simulations

To investigate the stability and flexibility of molecular complexes in a hydrated environment, studies using MD simulations were conducted. The simulations were performed using Gromacs 2019 software⁵⁹ and chemistry at Harvard macromolecular mechanics (CHARMM36) force field.⁶⁰,⁶¹ Ligand topology files were generated using the CHARMM general force field server (CGenFF).⁶²,⁶³ The associated penalty scores were evaluated to assess the reliability of the assigned parameters. For all lead compounds, the majority of parameters presented low penalty values (< 10), while a limited number fell within the moderate range (10-50), which is considered acceptable for exploratory molecular dynamics simulations. Importantly, no parameters exceeded the critical threshold (> 50), indicating that the generated topologies are suitable for reliable MD analysis without the need for further quantum mechanical refinement. A detailed summary of the penalty scores for the principal compounds is provided in Table S2 (SI section). The protein and protein-ligand complexes obtained from molecular docking were solvated in a cubic box with transferable intermolecular potential with 3 points (TIP3P) water molecules, which employ a three-point transferable intermolecular potential. To neutralize the net charge of the system, counter-ions were added, and the ionic strength was adjusted to 0.150 mol L^-1 NaCl, mimicking physiological conditions. In each system, we conducted a minimization of 50,000 steps, followed by a 125 ps constant-number-of-particles, constant-volume, and constant-temperature (NVT) equilibration step with a timestep of 1 fs. Subsequently, a 100 ns NPT simulation at 300 K and 1 bar was performed using the Berendsen method,⁶¹ with coupling every 2 ps and a timestep of 2 fs. Short-range interactions such as Lennard-Jones and Coulombic interactions were calculated with a cutoff radius of 12 Å, and long-range electrostatic interactions were treated using the particle mesh ewald (PME) method.⁶² Although the Berendsen barostat⁶³ is not the most rigorous method for pressure control, it was employed here due to its numerical stability, lower cost, and widespread use in exploratory biomolecular simulations. This choice does not compromise the comparative analysis between the ligands, which is the main focus of this study. The system temperature was maintained constant using a modified Berendsen thermostat.⁶⁴ From the generated trajectory, GROningen MAchine for Chemical Simulations (GROMACS) was employed to perform structural and energy analyses, such as root mean square deviation (RMSD), root mean square fluctuation (RMSF), and radius of gyration (RoG), principal component analysis (PCA), binding-free energy and MD-ligand-receptor 2023 developed by Pieroni et al.⁶⁵ to study time-dependent interactions.

Binding free energy calculations using MM/PBSA

To determine the binding free energy in complexes, we employed the MM/PBSA approach,⁶⁶ utilizing an ensemble from MD simulations. The trajectory starting from the stable equilibrium state was used to estimate the binding free energy in protein-ligand complexes. In essence, it was utilized the equation below to calculate the binding free energy (ΔG_bind) between the unbound protein (G_pro) and the ligand (G_lig), based on the free energy of the protein-ligand complex (G_plc):

(1) $\Delta {\text{G}}_{\text{bind}}={\text{G}}_{\text{plc}}-\left({\text{G}}_{\text{pro}}+{\text{G}}_{\text{lig}}\right)$

The MM/PBSA method enables the calculation of the energy associated with the binding of a ligand to a protein, incorporating various factors such as solvation energies. This calculation is performed through a complete thermodynamic cycle, considering both the potential energy and the energy related to polar and nonpolar solvation. In this study, we utilized MM/PBSA to predict the binding free energy, analyzing snapshots obtained throughout the equilibrium state.

Results and Discussion

The selection of the dopamine D(2) receptor as the molecular target in this study was based on three considerations. First, this receptor is a well-established pharmacological target in the treatment of several neurological and psychiatric disorders, including schizophrenia and Parkinson’s disease, where modulation of D(2) activity (agonism or antagonism) plays a central therapeutic role. This makes it a highly relevant target for the identification of new bioactive compounds. Second, ligand-based target prediction approaches, including the SEA and SwissTargetPrediction,²⁵,²⁹ consistently identified the D(2) receptor as the top-ranked biological target for the investigated alkaloids, based on structural similarity to known dopaminergic ligands. At last, the availability of a high-resolution crystal structure (PDB ID: 6CM4), co-crystallized with a clinically relevant ligand (risperidone), enables accurate structure-based modeling. The presence of a well-defined orthosteric binding site and experimentally validated key residues (e.g., Asp114, Tyr408) supports reliable docking and molecular dynamics simulations.

The dopamine D(2) receptor protein was modeled using the structure retrieved from the protein data bank, with PDB ID: 6CM4, at a resolution of 2.87 Å and a free R-value of 0.249 Å. This protein consists of a single chain and 423 residues. The structure was refined, removing small compounds and water molecules, using the Schrodinger maestro 2023-2 tool. The refinement process is illustrated in Figures S6 and S7 (SI section), along with the corresponding Ramachandran plot performed with PROCHECK (University College London (UCL), 2024), providing information on the stereochemistry of the target protein. Analysis of the plot revealed that 92.6% of residues are in favored regions, 7.4% in additionally allowed regions, and 0.0% in generously allowed regions. The predominance of residues in favored regions suggests a high-quality structure, with an overall score of 90%.

The integrated computational workflow employed in this study enabled the identification of alkaloids with favorable interaction profiles toward the dopamine D(2) receptor. Initial ADMET screening indicated that a subset of compounds exhibited suitable pharmacokinetic and toxicity profiles, justifying their progression to structure-based analyses.

From the application of the scoring method that combines ADMET properties, alkaloids were ranked as presented in Table 1. Among these, those showing the three highest values were selected as lead compounds, namely pallidine, N-methylcoclaurine, juziphine, orientaline, reticuline, 13-hydroxy-discretinine, discretamine (in descending order of value).

Table 1.
Ranking of alkaloid scores (0-1 scale)

Type	Name	Drug-score
Morphinanedienone	pallidinea	0.89
Benzyltetrahydroisoquinoline	N-methylcoclaurinea	0.88
	juziphinea	0.88
	orientalinea	0.87
	reticulinea	0.87
Tetrahydroprotoberbrine	13-hydroxy-discretininea	0.87
	discretaminea	0.87
	13-hydroxy-2,3,9,10-tetramethoxyprotoberberine	0.85
	coreximine	0.85
	isocoreximine	0.85
	stepholidine	0.85
Benzyltetrahydroisoquinoline	magnococline	0.84
Tetrahydroprotoberberine	discretine	0.83
	corytenchine	0.83
	xylopinine	0.82
	corypalmine	0.67
Benzyltetrahydroisoquinoline	coclaurine	0.50
Tetrahydroprotoberberine	7,8-dihydroxypalmatine	0.48

• ^a
  Top three scores were selected as lead compounds for further detailed analysis.

The main ADMET characteristics of the lead compounds were evaluated and are detailed in Table S1 (SI section), while the radar plot representation is shown in Figures S2 and S3 (SI section). The results obtained indicate that the compounds comply with the toxicity profile considered acceptable. Additionally, the bioavailability of the compounds, as evidenced by the radar plot, demonstrated favorable indicators.

DFT calculations

HOMO-LUMO analysis

The frontier molecular orbitals (HOMO-LUMO) provide essential insights into the chemical reactivity and kinetic stability of the studied alkaloids. Based on the energies of these orbitals (E_HOMO and E_LUMO), global reactivity descriptors, such as electronegativity (χ), chemical hardness (η), and electronic chemical potential (μ), were calculated following the theorem of Koopmans (equations 2-4).⁶⁷-⁷⁰

(2) $\chi =\frac{\left(\text{I}+\text{A}\right)}{2}$

(3) $\eta =\frac{\left(\text{I}-\text{A}\right)}{2}$

(4) $\mu =-\frac{\left(\text{I}+\text{A}\right)=-\chi }{2}$

where I = −E_HOMO and A = −E_LUMO is the ionization energy and electron affinity, respectively. Furthermore, the electrophilicity index (ω = µ²/2η), which measures the energy stabilization upon charge transfer, was used to classify the molecules as strong (ω > 1.5 eV), moderate (0.8 < ω < 1.5 eV), or marginal (ω < 0.8 eV)⁷¹ electrophiles. The energy HOMO-LUMO gap (E_g), acting as a measure of molecular hardness, was also evaluated, where larger values indicate greater electronic stability. The frontier orbital distributions and the calculated reactivity parameters are summarized in Table 2 and Figure 2, respectively.

Table 2.
Reactivity descriptors for 13-hyroxy-discretinine (13HD), discretamine (D), juziphine (J), N-methycoclaurine (N), orientaline (O), palidine (P) and reticuline (R)

Parameter	13 HD	D	J	N	O	P	R
Energy / a.u.	−1206.5	−1091.93	−978.69	−978.64	−1093.11	−1091.94	−1093.11
Dipole moment / Debye	5.5	3.26	4.4	2	3.53	6.41	0.93
EHOMO / eV	−5.3	−5.68	−5.14	−5.55	−5.39	−5.52	−5.45
ELUMO / eV	−0.02	0.04	0.09	0.09	−0.05	−1.6	0.1
EHOMO-LUMO / eV	5.28	5.72	5.23	5.64	5.34	3.92	5.55
Hardness (η) / eV	5.29	5.7	5.19	5.6	5.37	4.72	5.5
Chemical potential (μ) / eV	−2.66	−2.82	−2.53	−2.73	−2.72	−3.56	−2.68
Electronegativity (χ) / eV	2.66	2.82	2.53	2.73	2.72	3.56	2.68
Electrophilicity index (ω) / eV	0.67	0.7	0.61	0.67	0.69	1.34	0.65

Figure 2.
HOMO-LUMO orbitals and energy gap for (a) 13-hydroxy-discretinine (up) and discretamine (down), (b) juziphine (up) and N-methylcoclaurine (down), (c) orientaline (up) and palidine (down), (d) reticuline.

The electronic properties, such as E_g, are essential indicators of chemical stability and reactivity. When comparing alkaloids with similar scaffolds, the analysis reveals that 13-hydroxy-discretinine (13HD, E_g = 5.28 eV) is more reactive than discretamine (D, E_g = 5.72 eV) (Figure 2a). A similar trend is observed for juziphine (J, E_g = 5.23 eV) compared to N-methylcoclaurine (N, E_g = 5.64 eV) (Figure 2b). Notably, pallidine (P, E_g = 3.92 eV) stands out as significantly more reactive than orientaline (O, E_g = 5.34 eV) and reticuline (R, E_g = 5.55 eV) (Figures 2c and 2d). Regarding the electrophilicity index (ω), most compounds were classified as marginal electrophiles (ω < 0.8), with the exception of pallidine, which acts as a moderate electrophile (ω = 1.34, Table 2). In the context of the D(2) receptor, the relatively low energy gaps observed for pallidine and N-methylcoclaurine suggest a favorable chemical reactivity that facilitates induced-fit adjustments during the binding process. Furthermore, the HOMO density localized primarily on the aromatic rings indicates regions prone to donating electron density to the electrophilic residues of the receptor. This electronic profile, characterized by high chemical hardness (η > 4) and electronegativity (χ > 2), suggests that while these molecules are structurally stable, they possess the necessary electronic distribution to enhance interactions with the binding site of the receptor, stabilizing the negative charge and favoring non-covalent bonding with specific residues.

The electronic structure analysis complements the docking and MD results by providing insight into the intrinsic reactivity of the investigated alkaloids. Compounds such as pallidine, characterized by a reduced HOMO-LUMO energy gap, exhibited enhanced electronic adaptability, which may facilitate stabilizing interactions within the receptor environment during dynamic simulations.

Electrostatic potential maps

The molecular electrostatic potential (MEP) provides a 3D visualization of charge distribution, where red regions (negative potential) and blue regions (positive potential) represent nucleophilic and electrophilic sites, respectively. These maps are essential for characterizing atomic reactivity, molecular dipole moments, and electrostatic interactions in both molecular dynamics and quantum chemical calculations.⁶⁹-⁷³ The MEPs for the studied alkaloids are presented in Figure 3 and reveal a well-defined charge polarization pattern that directly underpins the interaction mechanism with the D(2) receptor. Across the alkaloids, regions of negative electrostatic potential are consistently localized on oxygen and nitrogen atoms, while positive regions are concentrated on hydrogen atoms, creating a dipolar framework that favors directional hydrogen bonding and electrostatic complementarity with key residues such as Asp114.

Figure 3.
Electrostatic potential map for (a) 13-hydroxy-discretinine, (b) discretamine, (c) juziphine, (d) N-methylcoclaurine, (e) orientaline, (f) palidine, (g) reticuline.

For 13HD (Figure 3a), negative potential is concentrated over O5 (label 21, −0.058 eV), O2 (label 11, −0.046 eV), O1 (label 3, −0.046 eV), N (label 15, −0.043 eV), O3 (label 8, −0.029 eV), and O4 (label 23, −0.023 eV), indicating favorable protonation sites. Positive potential is located over hydrogens H23 (label 45, 0.029 eV) and H13 (label 29, 0.043 eV). D (Figure 3b) exhibits negative regions at O2 (label 24, −0.064 eV), O4 and O3 (labels 23 and 21, −0.060 eV), and O1 (label 2, −0.056 eV), with positive sites at H21 (label 44, 0.065 eV) and H22 (label 45, 0.053 eV).

Regarding J (Figure 3c) and N (Figure 3d), negative potentials are centered on oxygens O1, O2, and O3 (ranging from −0.013 to −0.063 eV), while positive potentials reside on H14 and H22 (0.048 to 0.056 eV). Structure O (Figure 3e) shows negative potential over O1-O4 (from −0.057 to −0.067 eV) and positive potential at H9 (0.051 eV) and H21 (0.60 eV). P (Figure 3f) demonstrates the most intense negative potential at O2 (label 1, −0.086 eV) and O1 (−0.076 eV), with positive potential at H21 (0.044 eV). Finally, R (Figure 3g) presents negative potential at O1-O4 and N (ranging from −0.045 to −0.071 eV) and positive sites at H14 (0.057 eV) and H21 (0.052 eV). These regions of opposing potential facilitate molecular interactions in solution, such as dimer formation, and are crucial for stabilizing the ligands within the specific nucleophilic and electrophilic regions of the dopamine D(2) receptor active site.

Notably, compounds like pallidine exhibit more intense negative potential regions (e.g. O2 and O1), indicating a stronger affinity for electron donation and interaction with electrophilic or positively polarized receptor sites, which aligns with their higher interaction persistence observed in MD simulations.⁷⁴,⁷⁵ This suggests that electrostatic preorganization plays a predictive role in binding stability, where molecules with more pronounced and spatially accessible charge separation are better suited to form stable and long-lived interactions.

Molecular docking

Molecular docking results (Table 3) yielded negative binding energies for all complexes, confirming the spontaneity of alkaloid-receptor association. While benzyltetrahydroisoquinoline and morphinanedienone derivatives showed superior pharmacokinetic profiles, tetrahydroprotoberberine alkaloids, led by stepholidine (–9.16 kcal mol^-1), achieved higher docking scores. This is attributed to their multi-ring scaffolds, which facilitate extensive receptor contact, though slightly affecting their overall drug scores.

Table 3.
Binding energy obtained from docking all types of alkaloids studied with the dopamine D(2) receptor target protein (6CM4)

Type	Ligand	Binding energy / (kcal mol-1)
Benzyltetrahydroisoquinoline	magnococline	−8.55
	juziphine	−8.11
	reticuline	−8.04
	orientaline	−7.95
	N-methylcoclaurine	−7.61
	coclaurine	−7
Morphinanedienone	pallidine	−7.22
Tetrahydroprotoberberine	stepholidine	−9.16
	discretamine	−8.96
	isocoreximine	−8.73
	coreximine	−8.72
	corytenchine	−8.56
	discretine	−8.52
	corypalmine	−8.52
	7,8-dihydroxypalmatine	-8.44
	xylopinine	−8.27
	13-hydroxy-discretinine	−8.25
	13-hydroxy-2,3,9,10-tetramethoxyprotoberberine	−8.03

Specific interaction analyses (Figures 4 and 5) highlighted the role of key residues such as Asp114, Ser409, and His393. Pallidine and N-methylcoclaurine established 6 direct interactions each, involving crucial H-bonds with Ser409 and Asp114. Juziphine showed 5 interactions, including a critical H-bond with Asp114. Both orientaline and reticuline exhibited robust binding networks (9 and 8 interactions, respectively), primarily stabilized by multiple H-bonds with Ser409 and Asp114. The most extensive interaction profiles were observed for 13-hydroxy-discretinine (15 contacts) and discretamine (11 contacts), involving residues Trp413 and Ser409, although the latter presented a minor unfavorable interaction with Glu95 (Figure 5c)

Figure 4.
Binding sites of the labeled residues involved in interaction with the ligands obtained from molecular docking. (a) Palidine, (b) N-methylcoclaurine, (c) juziphine and (d) orientaline.

Figure 5.
Binding sites of the labeled residues involved in interaction with the ligands obtained from molecular docking. (a) Reticuline, (b) 13-hydroxy-discretine, (c) discretamine.

Molecular dynamics simulation

To deepen the analysis of the ligand-receptor complex stability, were performed MD simulations extended over 100 ns. From the obtained trajectories, we examined various detailed metrics, including the RMSD, RMSF, RoG, PCA, time-dependent interactions, and binding free energy using the MM/PBSA method.

Fluctuation and radius of gyration of complexes

The structural stability of the D(2) receptor complexes was assessed through RMSD analysis (Figure 6a and Table 4). Compared to the native system, complexes with N-methylcoclaurine and orientaline exhibited lower RMSD values, suggesting a higher degree of conformational conservation upon binding. In contrast, pallidine and juziphine induced more significant initial fluctuations before reaching equilibration. Notably, most systems presented lower standard deviations than the native complex, indicating that alkaloid binding generally enhances the structural rigidity of the receptor over the 100 ns trajectory. Reticuline was the exception, showing a less stable profile with a gradual increase in deviation until 30 ns.

Figure 6.
(a) Graph of the RMSD of the protein complexed with the investigated ligands compared to the native system over 100 ns, (b) RMSD of the lead compounds in contact, (c) RMSF plot of the native complex and complexes with lead compounds, zoomed in, it is observed the interaction domain with the ligands, where the hatched area represents the range of residue fluctuations corresponding to the native configuration, (d) RoG of complexes formed by the dopamine D(2) receptor protein with the investigated ligands compared to the native system.

Table 4.
Average value and standard deviation of RMSD, RoG, and RMSF for the complexes

Complexes	RMSD protein / nm		RMSD ligand / nm		RoG / nm		RMSF / nm
	AV	SD	AV	SD	AV	SD	AV	SD
Pallidine	0.73	0.14	0.58	0.07	2.93	0.06	0.32	0.1
N-Methyl coclaurine	0.57	0.18	0.81	0.31	2.95	0.04	0.34	0.14
Juziphine	0.72	0.11	0.71	0.15	2.89	0.04	0.31	0.11
Orientaline	0.56	0.1	0.37	0.09	2.9	0.04	0.31	0.11
Reticuline	0.86	0.26	0.57	0.08	3	0.04	0.42	0.17
13-Hydroxy-discretinine	0.74	0.14	0.79	0.17	2.96	0.04	0.43	0.17
Discretamine	0.74	0.15	0.79	0.31	2.71	0.08	0.32	0.12
Native	0.65	0.19	−	−	2.99	0.05	0.39	0.17

Regarding ligand stability within the orthosteric site (Figure 6b), pallidine, orientaline, and reticuline maintained the most consistent binding modes, characterized by low average RMSD and minimal standard deviations (Table 4). Conversely, N-methylcoclaurine and juziphine underwent visible conformational adjustments-reaching stability only after 25 ns and 62 ns, respectively. Discretamine exhibited a distinct multi-step stabilization process, reflecting its dynamic search for an optimal fit within the binding pocket during the simulation.

To identify the dynamic behavior of the most mobile residues and the effect of ligand binding on the flexibility of the target protein, the RMSF was investigated. As illustrated in the RMSF graph in Figure 6c, a more significant change in the flexibility of amino acids was observed in the ligand interaction region, specifically between residues 1030-1070. In this region, a notable reduction in flexibility was observed when the receptor was bound to discretamine, orientaline, juziphine, N-methylcoclaurine, and pallidine, and a slight relative increase in the presence of 13-hydroxy-discretinine and reticuline. This reduction suggests more stable interactions with the respective ligands, restricting the extensive movements of these residues. Except for the specific region (1030-1070), the systems exhibited similarity in residual fluctuations, with the average RMSF values of the complexes being reasonably close, as observed in Table 4. Higher RMSF values in residues indicated regions with greater fluctuation, possibly due to fewer or weaker interactions, while lower RMSF values reflected relatively rigid residues with more well-established interactions.

The RoG was determined to evaluate the compactness of the complexes formed by the dopamine D(2) receptor protein with the investigated ligands compared to the native complex. It is widely recognized that stably folded proteins exhibit smaller RoG values, reflecting greater compactness and dynamic stability. The RoG graph as a function of time (ns) in Figure 6d indicates that, overall, the receptor protein complexed with the lead alkaloids is slightly more compact than its native structure for most of the time. Additionally, the receptor bound to most of the investigated compounds showed lower average RoG values than in the case of the native complex, with the discretamine complex standing out for maintaining greater compactness throughout the entire simulation, as evidenced in Table 4. The RoG results indicate that the binding of alkaloids to the protein tends to promote the formation of more stable protein-ligand complexes.

Principal component analysis (PCA) and free energy landscape

To investigate the influence of alkaloid binding on the collective motions and conformational stability of the D(2) receptor, PCA was performed on the alpha-carbon (Cα) trajectories. By projecting the motion profiles onto the first two principal components (PC1 and PC2), we identified the essential dynamics that govern the protein-ligand stability (Figure 7).

Figure 7.
PCA constructed for the protein-ligand complex as a function of the best projections of the MD trajectory on the first (PC1) and second (PC2) eigenvectors, respectively.

The PCA scores reveal that complexes involving reticuline, juziphine, discretamine, pallidine, and orientaline occupied a significantly more confined conformational space compared to the native receptor. This reduction in motion amplitude suggests that these alkaloids induce a more cohesive and kinetically stable state in the protein. Conversely, complexes with 13-hydroxy-discretine and N-methylcoclaurine exhibited broader conformational coverage, indicating higher flexibility and less restrictive binding within the orthosteric site.

The thermodynamic stability and conformational distribution of the complexes were further evaluated through the free energy landscape (FEL), using PC1 and PC2 as reaction coordinates, using the following equation:

(5) $\Delta \text{F}=-{\text{k}}_{\text{B}}\text{T}\left(\text{ln}\left({\text{P}}_{\text{B}}\right)-ln\left({\text{P}}_{\text{A}}\right)\right)$

where ΔF represents the Gibbs free energy, k_B and T are the Boltzmann constant and temperature, respectively. P_B denotes the probability of the complex existing in a given conformation during dynamics, and P_A the probability of the most frequent conformation of the structure. The FEL maps the energy basins explored by the protein during the simulation, where deeper and more compact basins represent higher conformational stability (Figure 8).

Figure 8.
Free energy landscape of the ligand-bound complex of (a) pallidine, (b) N-methylcoclaurine, (c) juziphine, (d) orientaline, (e) reticuline, (f) 13-hydroxy-discretine, (g) discretamine, (h) native complex as a function of the projections along the first (PC1) and second (PC2) eigenvectors (in nm) obtained from MD trajectories. The colored bar represents the relative value of free energy in kcal mol^-1.

The FEL plots demonstrate that pallidine, orientaline, and discretamine form highly stable complexes, characterized by distinct and well-defined global minimum energy basins. These compact distributions, when compared to the native complex, suggest that these ligands effectively lock the receptor into stable, low-energy conformations. In contrast, N-methylcoclaurine, juziphine, and reticuline showed broader and more dispersed basins, reflecting greater conformational fluctuations and a more flexible binding mode. These results, combined with the PCA data, reinforce that pallidine and orientaline possess the most favorable structural and energetic profiles for D(2) receptor antagonism.

Temporal evolution of interactions

Ligand-receptor interactions were classified by bonding type and analyzed for their persistence throughout the 100 ns simulation. Significant disparities in interaction stability were observed between Group A (reticuline, pallidine, N-methylcoclaurine, and orientaline) and Group B (discretamine, juziphine, and 13-hydroxy-discretinine).

For group B, although a variety of bonds – including hydrogen bonds, hydrophobic interactions, π-stacking, and water bridges – were present, their durations were remarkably short. Specifically, hydrogen bonds maintained low average persistence: 23.4% for discretamine (max 43.3% at O4), 10.5% for juziphine (max 34.1% at O3), and 6.7% for 13-hydroxy-discretinine (max 19.3% at O5) (Figure 9). This transient behavior suggests a lower overall binding stability for these compounds.

Figure 9.
Types of bonds and interactions established by each individual atom of the lead compounds throughout the simulation period. In the bar graph, the atoms involved in the interactions are indicated on the x-axis, while the duration of these interactions during the simulation (in %) is represented on the y-axis in intervals. Different colors in the graph distinguish the various types of bonds or interactions observed.

In contrast, group A exhibited highly stable hydrogen bonds as the predominant interaction type. Reticuline demonstrated exceptional persistence, with the O1 atom maintaining contact for over 95% of the simulation. Similarly, substantial persistence was recorded for pallidine (ca. 72% at O1), N-methylcoclaurine (ca. 65% at O3), and orientaline (ca. 62% at O3) (Figure 9). These findings align with the RMSD analysis, confirming that group A ligands form significantly more robust complexes with the D(2) receptor compared to those in group B.

It was investigated the non-covalent interactions within the 6CM4 receptor, emphasizing residues ASP114, TYR408, TRP100, PHE389, PHE189, and PHE110 due to their high persistence across most complexes. This dynamic analysis refined docking predictions by revealing how interaction profiles evolve over time, as visualized in the heat maps of Figures S8-S10 (SI section), where intense horizontal lines indicate collective atomic interactions with specific residues.

Reticuline demonstrated robust stability through specific contacts between ASP114 and the O1 atom (95% persistence) and PHE389 with the C5 atom (68.1%). Notably, MD simulations showed that the docking-predicted interaction with the C6-C11 ring was not significant. For pallidine, TYR408 established long-duration interactions with the main ring (77.3%) and a moderate, unpredicted contact with atoms C6, C9, and C10 (50.4%). Furthermore, a predominant interaction between ILE403 and the O1 atom (65.5%) emerged during the simulation.

The stability of orientaline was primarily driven by the O3-ASP114 (67.4%) and TYR408-ring (50.7%) interactions. In contrast, N-methylcoclaurine favored a TRP100 interaction (65.9%) over the docking-predicted PHE389, while its contact with ASP114 was more persistent via the O3 atom (35.3%) than the predicted O1 atom. Compounds with lower persistence included discretamine (PHE389, 45.8%; PHE110, 34.6%) and 13-hydroxy-discretinine, which favored TRP100 (44.2%) and TYR408 (36.3%) instead of the docking-suggested ILE184. Finally, juziphine exhibited the lowest stability, with maximum persistence values below 17% for residues PHE189 and TYR408, diverging from its initial docking profile.

Analysis and decomposition of binding free energy using the MM/PBSA method

To evaluate the individual contributions to dopamine D(2) receptor inhibition, MM/PBSA energy decomposition was performed (Figure 10). The analysis revealed that while residues like Asp114 and HSD393 exhibited strong ligand interactions, their effective contribution to binding stability was often offset by unfavorable polar solvation effects (Figures 10a and 10b). Consequently, van der Waals interactions emerged as the primary driving force for complex stability (Figure 10c), with Tyr408, Phe189, Trp100, Ile184, and Phe389 identified as the key residues facilitating these interactions (Figure 10d).

Figure 10.
Decomposition of energy by residue: (a) contribution of electrostatic energy, (b) polar solvation energy, (c) van der Waals interactions, (d) total energy.

Figures S11-S13 (SI section) shows the chemical structure of the compounds and highlights the main residues responsible for direct binding to the receptor.

All lead compounds displayed favorable binding free energies (ΔG_Bind), as summarized in Figure 11. Van der Waals contributions (ΔG_vdW - EVDWAALS) were the most significant, ranging from −29.91 kcal mol^-1 (juziphine) to −45.4 kcal mol^-1 (reticuline). While electrostatic interactions (ΔG_EL - EEL) varied – being favorable only for pallidine and 13-hydroxy-discretinine – polar solvation energy (ΔG_PB - EPB) acted as an unfavorable penalty in nearly all complexes. Based on the total energetic profile, orientaline and discretamine were ranked as the most and least stable ligands within the 6CM4 active site, respectively, with the remaining alkaloids showing intermediate binding affinities. The binding free energy values obtained are comparable to those reported for classical dopamine D(2) receptor ligands in computational studies,⁷⁶ reinforcing their potential as alternative structures for dopaminergic modulation.

Figure 11.
Binding free energy based on MM/PBSA of the system formed by the lead compounds with the protein 6CM4.

MD, PCA and DFT-driven implications and insights from results

Molecular dynamics results reveal that ligand efficacy is not governed solely by initial docking scores, but by the ability to stabilize the D(2) receptor in low-energy conformational states. Complexes with pallidine and orientaline exhibited lower protein RMSD values (ca. 0.56-0.73 nm), reduced residue fluctuations in the binding region (1030-1070), and more compact structures (lower RoG), indicating that these ligands promote conformational restriction of the receptor. This effect is further supported by PCA and FEL analyses, which show confined conformational spaces and well-defined global minima, suggesting that these compounds effectively lock the receptor into stable states. Pharmacologically, such stabilization is consistent with antagonist-like behavior at the D(2) receptor, as many antipsychotic agents act by restricting receptor conformational flexibility. Additionally, the high persistence of key hydrogen bonds (particularly with Asp114 and Tyr408 with > 60-95%) combined with dominant van der Waals contributions, indicates that binding stability arises from a synergy between specific polar anchoring and hydrophobic packing, a hallmark of high-affinity GPCR ligands. On the other hand, DFT analysis provided a complementary mechanistic layer by linking electronic structure to binding behavior. Pallidine, which exhibits the lowest HOMO-LUMO gap (3.92 eV) and the highest electrophilicity index (ω = 1.34), demonstrates enhanced electronic adaptability, facilitating charge redistribution during ligand-receptor interaction. This property likely contributes to its ability to maintain persistent interactions during MD simulations and to stabilize key residues within the binding pocket. In contrast, compounds with larger energy gaps (> 5 eV) display reduced reactivity and correspondingly less stable dynamic profiles. The localization of HOMO density over aromatic rings and heteroatoms (N/O), together with MEP-derived nucleophilic regions, supports electron donation to electrophilic residues such as Asp114, reinforcing hydrogen bonding and electrostatic complementarity. From a pharmacological perspective, these findings suggest that moderate electrophilicity combined with strategic charge distribution is a key determinant of dopaminergic activity, providing a rational basis for structure-activity optimization. Specifically, tuning substituents to modulate electronic softness without compromising structural rigidity may enhance both binding affinity and selectivity for the D(2) receptor.

Conclusions

This study provides a comprehensive computational investigation into the interaction mechanisms between alkaloids derived from Annonaceae species and the dopamine D(2) receptor, integrating pharmacokinetic screening, molecular docking, electronic structure analysis, molecular dynamics simulations, and binding free energy calculations. The combined use of these complementary methodologies enabled a consistent and mechanistically grounded assessment of the stability, reactivity, and binding affinity of the investigated compounds.

Among the analyzed alkaloids, orientaline, pallidine and N-methylcoclaurine emerged as the most promising candidates, exhibiting a favorable balance between electronic adaptability, structural complementarity, and persistent intermolecular interactions within the D(2) receptor binding pocket. These compounds displayed reduced conformational fluctuations during long-timescale molecular dynamics simulations, well-defined free energy minima, and the most favorable MM/PBSA binding free energy profiles, predominantly driven by van der Waals interactions complemented by stable hydrogen bonding with key receptor residues.

Electronic structure calculations further supported these findings by revealing that the lead compounds possess suitable global reactivity descriptors, suggesting sufficient chemical stability combined with the electronic flexibility required for effective receptor engagement. The convergence between DFT-derived descriptors and dynamic stability highlights the relevance of correlating intrinsic molecular properties with receptor-level behavior in the rational evaluation of dopaminergic ligands. The insights obtained herein support the further experimental validation of the identified lead compounds through in vitro binding assays and functional studies, as well as future structure-activity relationship investigations aimed at optimizing selectivity and efficacy. Although the present results are based exclusively on computational approaches, they provide a solid molecular-level framework for the identification of alkaloid scaffolds with potential dopaminergic activity. As a future experimental validation strategy, the most promising alkaloid-target complexes predicted by docking and molecular dynamics simulations could be evaluated using surface plasmon resonance (SPR) or isothermal titration calorimetry (ITC).⁷⁷,⁷⁸ These techniques would allow direct determination of binding affinity and, in the case of ITC, thermodynamic parameters associated with complex formation, thereby providing experimental support for the predicted ligand-protein interactions. Overall, this work reinforces the value of natural-product-inspired scaffolds in central nervous system (CNS) drug discovery and demonstrates the effectiveness of integrated in silico strategies for the early-stage identification of novel dopamine D(2) receptor modulators.

Supplementary Information

Supplementary information (ADMET characteristics tables, and additional figures) is available free of charge at http://jbcs.sbq.org.br as PDF file.

Supplementary PDF

Supplementary PDF

Data Availability Statement

Data will be made available on request.

Acknowledgments

The authors used Gemini and ChatGPT to assist in editing the text and improving style. All suggestions generated by the AI were subsequently reviewed and edited by the authors as needed.

The authors thanks FAPEAM for financial support, UFAM for infrastructure and computational resources offered by the Laboratory of Theoretical and Computational Chemistry (LQTC) and the Department of Chemistry (DQ), FUCAPI, CNPq (Grant No. 306335/2023-9), and FitoAmazônia Research Network (Pró-Amazônia/CNPq Grant No. 445615/2024-9).

References

• ¹ Almeida, C. V. D.; Yara, R.; Almeida, M. D.; Pesq. Agropec. Bras. 2005, 40, 467. [Crossref]
  » [Crossref]
• ² Pereira, M. D. M.; Jácome, R. L. R. P.; Alcântara, A. F. D. C.; Alves, R. B.; Raslan, D. S.; Quim. Nova 2007, 30, 970. [Crossref]
  » [Crossref]
• ³ Newman, D. J.; Cragg, G. M.; Snader, K. M.; J. Nat. Prod. 2003, 66, 1022. [Crossref]
  » [Crossref]
• ⁴ Ramadinha, R. R.; Teixeira, R. D. S.; Bomfim, P. C.; Mascarenhas, M. B.; França, T. N.; Peixoto, T. C.; Costa, S. Z. R.; Peixoto, P. V.; Rev. Bras. Med. Vet. 2016, 38, 65. [Link] accessed in July 2026
  » [Link]
• ⁵ Scazzocchio, F.; Cometa, M. F.; Tomassini, L.; Palmery, M.; Planta Med. 2001, 67, 561. [Crossref]
  » [Crossref]
• ⁶ Biavatt, M. W.: Estudos sobre 8,10-di-n-Propil-Lobelidiol: Um Novo Alcaloide Isolado de Siphocampylus Verticillatus (Cham.) G.Don., Campanulaceae; MSc Dissertation, Universidade Federal de Santa Catarina, Florianópolis, Brazil, 1994. [Link] accessed in June 2026
  » [Link]
• ⁷ Corrêa, M. P.; Dicionário das Plantas Úteis do Brasil e das Exóticas Cultivadas; IBDF Ministério da Agricultura: Rio de Janeiro, Brazil, 1984.
• ⁸ Couvreur, T. L. P.; Pirie, M. D.; Chatrou, L. W.; Saunders, R. M. K.; Su, Y. C. F.; Richardson, J. E.; Erkens, R. H. J.; J. Biogeogr. 2011, 38, 664. [Crossref]
  » [Crossref]
• ⁹ Leboeuf, M.; Cavé, A.; Bhaumik, P. K.; Mukherjee, B.; Mukherjee, R.; Phytochemistry 1980, 21, 2783. [Crossref]
  » [Crossref]
• ¹⁰ Rupprecht, J. K.; Hui, Y.-H.; McLaughlin, J. L.; J. Nat. Prod. 1990, 53, 237. [Crossref]
  » [Crossref]
• ¹¹ Lúcio, A. S. S. C.; Almeida, J. R. G. S.; Cunha, E. V. L.; Tavares, J. F.; Barbosa Filho, J. M. In The Alkaloids: Chemistry and Biology; Cordell, G. A., ed.; Elsevier: Amsterdam, Netherlands, 2015, p. 233-409.
• ¹² de Souza, C.; Nardelli, V.; Paz, W.; Pinheiro, M.; Rodrigues, A.; Bomfim, L.; Soares, M.; Bezerra, D.; Chaar, J.; Koolen, H.; da Silva, F.; Costa, E. V.; Quim. Nova 2020, 43, 1397. [Crossref]
  » [Crossref]
• ¹³ Araújo, M. S.; Silva, F. M. A.; Koolen, H. H. F.; Costa, E. V.; Biochem. Syst. Ecol. 2020, 92, 104105. [Crossref]
  » [Crossref]
• ¹⁴ Ngouonpe, A. W.; Mbobda, A. S. W.; Happi, G. M.; Mbiantcha, M.; Tatuedom, O. K.; Ali, M. S.; Lateef, M.; Tchouankeu, J. C.; Kouam, S. F.; Biochem. Syst. Ecol. 2019, 83, 22. [Crossref]
  » [Crossref]
• ¹⁵ Santos, R. C.; Souza, A. V.; Silva, M. A.; Cruz, K. C. V.; Kassuya, C. A. L.; Cardoso, C. A. L.; Vieira, M. D. C.; Formagio, A. S. N.; J. Ethnopharmacol. 2018, 211, 9. [Crossref]
  » [Crossref]
• ¹⁶ Costa, E. V.; Pinheiro, M. L. B.; Souza, A. D. L. D.; Barison, A.; Campos, F. R.; Valdez, R. H.; Ueda-Nakamura, T.; Filho, B. P. D.; Nakamura, C. V.; Molecules 2011, 16, 9714. [Crossref]
  » [Crossref]
• ¹⁷ Silva, D. B.; Tulli, E. C. O.; Militão, G. C. G.; Costa-Lotufo, L. V.; Pessoa, C.; Moraes, M. O.; Albuquerque, S.; Siqueira, J. M.; Phytomedicine 2009, 16, 1059. [Crossref]
  » [Crossref]
• ¹⁸ Costa, E. V.; Pinheiro, M. L. B.; Xavier, C. M.; Silva, J. R. A.; Amaral, A. C. F.; Souza, A. D. L.; Barison, A.; Campos, F. R.; Ferreira, A. G.; Machado, G. M. C.; Leon, L. L. P.; J. Nat. Prod. 2006, 69, 292. [Crossref]
  » [Crossref]
• ¹⁹ Perez, E.; Saez, J.; Blair, S.; Franck, X.; Figadere, B.; Lett. Org. Chem. 2004, 1, 102. [Crossref]
  » [Crossref]
• ²⁰ Costa, E. V.; Teixeira, S. D.; Marques, F. A.; Duarte, M. C. T.; Delarmelina, C.; Pinheiro, M. L. B.; Trigo, J. R.; Maia, B. H. L. N. S.; Phytochemistry 2008, 69, 1895. [Crossref]
  » [Crossref]
• ²¹ Costa, E. V.; Cruz, P. E. O.; Lourenço, C. C.; Moraes, V. R. S.; Nogueira, P. C. L.; Salvador, M. J.; Nat. Prod. Res. 2013, 27, 1002. [Crossref]
  » [Crossref]
• ²² Costa, R. A.; Barros, G. D. A.; Silva, J. N.; Oliveira, K. M. T.; Bezerra, D. P.; Soares, M. B. P.; Costa, E. V.; J. Mol. Struct. 2021, 1229, 129844. [Crossref]
  » [Crossref]
• ²³ Quayum, S. T.; Esha, N. J. I.; Siraji, S.; Abbad, S. S. A.; Alsunaidi, Z. H. A.; Almatarneh, M. H.; Rahman, S.; Alodhayb, A. N.; Alibrahim, K. A.; Kawsar, S. M. A.; Uddin, K. M.; MethodsX 2024, 12, 102537. [Crossref]
  » [Crossref]
• ²⁴ Saha, S.; Prinsa; Kawsar, S. M. A.; Chem. Biodiversity 2025, 22, e202402521. [Crossref]
  » [Crossref]
• ²⁵ Daina, A.; Michielin, O.; Zoete, V.; J. Chem. Inf. Model. 2014, 54, 3284. [Crossref]
  » [Crossref]
• ²⁶ Daina, A.; Zoete, V.; ChemMedChem 2016, 11, 1117. [Crossref]
  » [Crossref]
• ²⁷ Daina, A.; Michielin, O.; Zoete, V.; Sci. Rep. 2017, 7, 42717. [Crossref]
  » [Crossref]
• ²⁸ Pires, D. E. V.; Blundell, T. L.; Ascher, D. B.; J. Med. Chem. 2015, 58, 4066. [Crossref]
  » [Crossref]
• ²⁹ Banerjee, P.; Eckert, A. O.; Schrey, A. K.; Preissner, R.; Nucleic Acids Res. 2018, 46, W257. [Crossref]
  » [Crossref]
• ³⁰ Sander, T.; Freyss, J.; von Korff, M.; Reich, J. R.; Rufener, C.; J. Chem. Inf. Model. 2009, 49, 232. [Crossref]
  » [Crossref]
• ³¹ Keiser, M. J.; Roth, B. L.; Armbruster, B. N.; Ernsberger, P.; Irwin, J. J.; Shoichet, B. K.; Nat. Biotechnol. 2007, 25, 197. [Crossref]
  » [Crossref]
• ³² Sotriffer, C., ed.; Virtual Screening: Principles, Challenges, and Practical Guidelines, 1^st ed.; Wiley-VCH: Weinheim, Germany, 2011.
• ³³ Eckert, H.; Bajorath, J.; Drug Discovery Today 2007, 12, 225. [Crossref]
  » [Crossref]
• ³⁴ Muegge, I.; Mukherjee, P.; Expert Opin. Drug Discovery 2016, 11, 137. [Crossref]
  » [Crossref]
• ³⁵ Friedman, J. I.; Temporini, H.; Davis, K. L.; Biol. Psychiatry 1999, 45, 1. [Crossref]
  » [Crossref]
• ³⁶ Millan, M. J.; J. Pharmacol. Exp. Ther. 2000, 295, 853. [Crossref]
  » [Crossref]
• ³⁷ Glennon, J.; Wadman, W.; McCreary, A.; Werkman, T.; CNS Neurol. Disord.: Drug Targets 2006, 5, 3. [Crossref]
  » [Crossref]
• ³⁸ Martel, J. C.; McArthur, S. G.; Front. Pharmacol. 2020, 11, 1003. [Crossref]
  » [Crossref]
• ³⁹ Cho, H. S.; Baek, D. J.; Baek, S. S.; J. Exercise Nutr. Biochem. 2014, 18, 379. [Crossref]
  » [Crossref]
• ⁴⁰ Fan, X.; Xu, M.; Hess, E. J.; Neurobiol. Dis. 2010, 37, 228. [Crossref]
  » [Crossref]
• ⁴¹ Mehta, M.; Riedel, W.; Curr. Pharm. Des. 2006, 12, 2487. [Crossref]
  » [Crossref]
• ⁴² Liggins, J.; McGill Sci. Undergrad. Res. J. 2009, 4, 39. [Crossref]
  » [Crossref]
• ⁴³ Pineda, P. S.; Delaveau, P.; Blin, O.; Nieoullon, A.; Clin. Neuropharmacol. 2005, 28, 228. [Crossref]
  » [Crossref]
• ⁴⁴ Van der Schyf, C. J.; Expert Rev. Clin. Pharmacol. 2011, 4, 293. [Crossref]
  » [Crossref]
• ⁴⁵ Kaasinen, V.; Vahlberg, T.; Stoessl, A. J.; Strafella, A. P.; Antonini, A.; Mov. Disord. 2021, 36, 1781. [Crossref]
  » [Crossref]
• ⁴⁶ Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew, R. K.; Goodsell, D. S.; Olson, A. J.; J. Comput. Chem. 2009, 30, 2785. [Crossref]
  » [Crossref]
• ⁴⁷ Wang, S.; Che, T.; Levit, A.; Shoichet, B. K.; Wacker, D.; Roth, B. L.; Nature 2018, 555, 269. [Crossref]
  » [Crossref]
• ⁴⁸ Burley, S. K.; Berman, L. M.; Bhikadiya, C.; Bi, C.; Chen, L.; Costanzo, L. D.; Christie, C.; Dalenberg, K.; Duarte, J. M.; Dutta, S.; Feng, Z.; Ghosh, S.; Goodsell, D. S.; Green, R. K.; Guranović, V.; Guzenko, D.; Hudson, B. P.; Kalro, T.; Liang, Y.; Lowe, R.; Namkoong, H.; Peisach, E.; Periskova, I.; Prlić, A.; Randle, C.; Rose, A.; Rose, P.; Sala, R.; Sekharan, M.; Shao, C.; Tan, L.; Tao, Y.-P.; Valasatava, Y.; Voigt, M.; Westbrook, J.; Woo, J.; Yang, H.; Young, J.; Zhuravleva, M.; Zardecki, C.; Nucleic Acids Res. 2019, 47, D464. [Crossref]
  » [Crossref]
• ⁴⁹ Schrödinger, LLC; Schrödinger Release 2023-1: Release Notes [Link] accessed in June 2026
  » [Link]
• ⁵⁰ Eberhardt, J.; Santos-Martins, D.; Tillack, A. F.; Forli, S.; J. Chem. Inf. Model. 2021, 61, 3891. [Crossref]
  » [Crossref]
• ⁵¹ O’Boyle, N. M.; Banck, M.; James, C. A.; Morley, C.; Vandermeersch, T.; Hutchison, G. R.; J. Cheminf. 2011, 3, 33. [Crossref]
  » [Crossref]
• ⁵² Jiménez, J.; Škalič, M.; Martínez-Rosell, G.; De Fabritiis, G.; J. Chem. Inf. Model. 2018, 58, 287. [Crossref]
  » [Crossref]
• ⁵³ Hu, J.; Liu, Z.; Yu, D.-J.; Zhang, Y.; Bioinformatics 2018, 34, 2209. [Crossref]
  » [Crossref]
• ⁵⁴ Ramírez, D.; Caballero, J.; Molecules 2018, 23, 1038. [Crossref]
  » [Crossref]
• ⁵⁵ Mahgoub, R. E.; Atatreh, N.; Ghattas, M. A.; Annu. Rep. Med. Chem. 2022, 59, 99. [Crossref]
  » [Crossref]
• ⁵⁶ Gaussview 5.0, Revision D.01; Gaussian, Inc.; Wallingford, CT, USA, 2003.
• ⁵⁷ Gaussian 09, Revision D.01; Gaussian, Inc.; Wallingford, CT, USA, 2009.
• ⁵⁸ Tirado-Rives, J.; Jorgensen, W. L.; J. Chem. Theory Comput. 2008, 4, 297. [Crossref]
  » [Crossref]
• ⁵⁹ Abraham, M. J.; Murtola, T.; Schulz, R.; Páll, S.; Smith, J. C.; Hess, B.; Lindahl, E.; SoftwareX 2015, 1-2, 19. [Crossref]
  » [Crossref]
• ⁶⁰ Huang, J.; MacKerell, A. D.; J. Comput. Chem. 2013, 34, 2135. [Crossref]
  » [Crossref]
• ⁶¹ Berendsen, H. J. C.; Postma, J. P. M.; Van Gunsteren, W. F.; DiNola, A.; Haak, J. R.; J. Chem. Phys. 1984, 81, 3684. [Crossref]
  » [Crossref]
• ⁶² Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G.; J. Chem. Phys. 1995, 103, 8577. [Crossref]
  » [Crossref]
• ⁶³ Bakker, M. J.; Sørensen, H. V.; Skepö, M.; J. Chem. Theory Comput. 2024, 20, 1423. [Crossref]
  » [Crossref]
• ⁶⁴ Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J. G. E. M.; J. Comput. Chem. 1997, 18, 1463. [Crossref]
  » [Crossref]
• ⁶⁵ Pieroni, M.; Madeddu, F.; Di Martino, J.; Arcieri, M.; Parisi, V.; Bottoni, P.; Castrignanò, T.; Int. J. Mol. Sci. 2023, 24, 11671. [Crossref]
  » [Crossref]
• ⁶⁶ Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S.; Chong, L.; Lee, M.; Lee, T.; Duan, Y.; Wang, W.; Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D. A.; Cheatham, T. E.; Acc. Chem. Res. 2000, 33, 889. [Crossref]
  » [Crossref]
• ⁶⁷ Mulliken, R. S.; J. Chem. Phys. 1934, 2, 782. [Crossref]
  » [Crossref]
• ⁶⁸ Janak, J. F.; Phys. Rev. B 1978, 18, 7165. [Crossref]
  » [Crossref]
• ⁶⁹ Pearson, R. G.; J. Am. Chem. Soc. 1963, 85, 3533. [Crossref]
  » [Crossref]
• ⁷⁰ Perdew, J. P.; Parr, R. G.; Levy, M.; Balduz, J. L.; Phys. Rev. Lett. 1982, 49, 1691. [Crossref]
  » [Crossref]
• ⁷¹ Pearson, R. G.; Coord. Chem. Rev. 1990, 100, 403. [Crossref]
  » [Crossref]
• ⁷² Bowen, J. P.; Allinger, N. L. In Reviews in Computational Chemistry; Lipkowitz, K. B.; Boyd, D. B., eds.; Wiley-VCH: New York, NY, USA, 1991, ch. 3.
• ⁷³ McCammon, J. A.; Harvey, S. C.; Dynamics of Proteins and Nucleic Acids, 1^st ed.; Cambridge University Press: Cambridge, UK, 1987.
• ⁷⁴ Tasi, G.; Pálinkó, I. In Molecular Similarity II; Sen, K. D., ed.; Springer: Berlin, Germany, 1995, ch. 3.
• ⁷⁵ Montagnani, R.; Tomasi, J.; J. Mol. Struct.: THEOCHEM 1993, 279, 131. [Crossref]
  » [Crossref]
• ⁷⁶ Bueschbell, B.; Barreto, C. A. V.; Preto, A. J.; Schiedel, A. C.; Moreira, I. S.; Molecules 2019, 24, 1196. [Crossref]
  » [Crossref]
• ⁷⁷ Du, X.; Li, Y.; Xia, Y.-L.; Ai, S.-M.; Liang, J.; Sang, P.; Ji, X.-L.; Liu, S.-Q.; Int. J. Mol. Sci. 2016, 17, 144. [Crossref]
  » [Crossref]
• ⁷⁸ Upadhyay, V.; Lucas, A.; Patrick, C.; Mallela, K. M. G.; Methods 2024, 225, 52. [Crossref]
  » [Crossref]

Edited by

• Editor handled this article:
  Paulo Augusto Netz (Associate)

Publication Dates

• Publication in this collection
  03 Aug 2026
• Date of issue
  2026

History

• Accepted
  23 Mar 2026