Evaluation of the Inhibitory Effect of The Enzyme Monoamine Oxidase A by Species of the Genus Spondias: An in silico and in vitro Study

Abstract

This study explores potential monoamine oxidase A (MAO-A) inhibitors from secondary metabolites of Spondias species for the treatment of depression. Using molecular docking, the interaction affinity and binding mode of the compounds with the active site of MAO-A were analyzed. Of the 104 metabolites identified, 78 showed significant interactions, with high affinity and specificity. Evaluation of the drug-like properties, according to Lipinski’s Rule of Five (molecular mass, hydrogen bond donors and acceptors and log P coefficient), showed that the majority met the criteria, suggesting good potential as drug candidates. Toxicity predictions were also carried out to ensure the safety of the compounds. In vitro MAO-A inhibition tests with ethanolic extracts of the species with the best docking performance indicated that Spondias metabolites are promising for new treatments for depression.

Keywords:
MAO-A inhibitors; secondary metabolites; depression treatment

Introduction

The use of plants for medicinal purposes is one of the oldest practices of humanity for the treatment of various ailments.¹ These practices often guide the search for bioactive compounds in plants, although there is not necessarily a direct disclosure of new therapeutic substances motivated exclusively by the popular use of medicinal plants. However, it is important to note that these practices contribute significantly to exploring the therapeutic potential of plants, many of which contain as yet unknown secondary metabolites.²

The secondary metabolites identified and/or isolated from plants are classes of compounds produced by different biosynthetic pathways, whose quantitative and qualitative variability can be influenced by a variety of factors, including the presence of natural predators. The production of these compounds can be modified according to interactions between physiological, ecological and biochemical processes.³,⁴

Thus, over the centuries, the various secondary metabolites present in plants have contributed as therapeutic resources in folk medicine, being used in the prevention, cure and treatment of various illnesses.⁵,⁶ In the practice of folk medicine, secondary metabolites are not used directly, but rather specific parts or combinations of different parts of plants, as transmitted orally between communities.

In this context, it can be seen that the chemical components produced by plants, which give them therapeutic activity, are called natural active ingredients. These substances have a variety of molecular structures with numerous functional groups in their carbon chains, and are organized into different classes according to their chemical similarities.⁶ For this reason, there has been an increasing number of studies into the use of secondary metabolites for pharmacological purposes, such as antitumor agents, antioxidants, anti-inflammatories and antimicrobials, among others.⁷,⁸

Nowadays, there is greater knowledge of the diversity of tools that science uses to study the effects of various substances on the body, as well as to make it possible to obtain new compounds from molecular prototypes.⁹ Thus, computational resources have emerged as an opportunity to carry out in silico experiments, reproducing the biological environment.

From this perspective, one of the most widely used in silico methods is molecular docking, which uses theoretical data for the docking between a protein and a ligand, with the aim of studying their interactions using structure-activity relationship methods, i.e., the molecular complex formed allows an understanding of the ligand and the protein, its specificities as a complex and energetic interactions, as well as considering electronic affinity characteristics and intermolecular interactions.¹⁰ In this way, this process selects possible ligands that act according to certain selection criteria. Subsequently, these ligands can be tested in vitro and in vivo to assess their activity, which can vary from inhibitors to agonists, or even inverse agonists, depending on the specific context.¹¹

Every day there is more research aimed at discovering new drug prototypes, which can be developed through the synthesis of new molecules designed to act on specific targets that are already known. It is also possible to explore existing molecules to identify their interactions with targets that are not yet known.¹⁰ This study, in turn, highlighted and studied the secondary metabolites that have been identified and/or isolated from the Spondias genus as monoamine oxidase (MAO) inhibitors.

The genus Spondias (belonging to the Anacardiaceae family) comprises between 14 and 21 species that are distributed worldwide. However, in Brazil, the species that stand out from this genus are the cajazeira (S. mombin), the umbuzeiro (S. tuberosa Arruda), the seriguela (S. purpurea), cajá-umbu (S. mombin × S. tuberosa) and cajarana or cajá-manga (S. dulcis). These species are exploited for their economic importance and also for their pharmaceutical properties.¹²,¹³ In the literature, for example, there are studies showing the use of species belonging to the genus Spondias in popular use. For example, species of the Spondias genus are used to treat infectious disorders (influenza, skin infections, viral hepatitis type B and C, respiratory diseases and others) in different regions of the world in traditional medicine.¹³

MAO is a group of isoenzymes found in the outer membrane of mitochondria, mainly in nerve, glial and other cells.¹⁴ MAO has two isoforms, monoamine oxidase A (MAO-A) and monoamine oxidase B (MAO-B),¹⁵ differing in substrate, sensitivity to inhibitors and amino acid sequence. MAO-A inhibition, the focus of this study, is used as a strategy in the treatment of depression, as studies have shown that MAO-A is closely related to this disorder, both due to its substrate specificity and the increase in enzyme levels. Depression is a multi-causal psychological disorder that can appear in anyone, regardless of age, gender, ethnicity or socioeconomic status, causing a range of emotions and disabling effects.¹⁶-¹⁸

On the other hand, MAO-B inhibition is being studied due to its association with Parkinson’s disease and Alzheimer’s disease, since studies¹⁷ show that this enzyme also performed altered activity in these pathologies. In addition, selective MAO-B inhibitors, such as IMAO-B, are used to treat Parkinson’s disease.

Furthermore, the prediction of selective inhibitors for MAO-A and MAO-B through computational chemistry makes it possible to find new drug candidates with high affinity and specificity for the target site, as well as with greater intrinsic activity, which can enhance the development of more efficient and selective drugs in order to minimize the effects of the hyperactivity of these enzymes in the pathological process.¹⁷-¹⁹

Species of the genus Spondias are extremely well known in traditional medicine for their therapeutic properties, being used to treat various situations, such as gastrointestinal disorders and infections. However, recent studies²⁰ have explored the potential of plant secondary metabolites as inhibitors of enzymes, including MAO-A, which plays a crucial role in neurotransmitter metabolism and is associated with the treatment of depression.

The choice to investigate species of the genus Spondias to find MAO-A inhibitors is justified for several reasons. First, the diversity of bioactive compounds present in these plants offers unexplored potential for the discovery of new drugs. Furthermore, Melfi et al.²⁰ highlighted the effectiveness of natural compounds such as MAO-B inhibitors, indicating that nature can be a rich and inspired source for the development of MAO inhibitors. As described by Melfi et al.,²⁰ the chemical modification of natural compounds has shown advances in the prevention of MAO-B, which leads us to explore Spondias compounds to inhibit MAO-A.

Furthermore, the choice to investigate MAO-A inhibitors in the current context is especially relevant due to the growing number of people suffering from mental disorders such as depression. The search for new, more effective treatments with fewer side effects is essential to meet this growing demand.

This innovative approach seeks not only to validate the therapeutic potential of species of the genus Spondias, but also to contribute to the development of new treatment alternatives for depression, a condition that affects millions of people globally.

The aim of this work is therefore to identify potential MAO-A inhibitors for the treatment of depression, using compounds identified and/or isolated from species of the genus Spondias, based on an analysis of the results of molecular docking calculations between the MAO-A enzyme and the compounds studied and, subsequently, to validate them by carrying out in vitro inhibition tests on the enzyme under study using extracts of the species with the best results in the molecular docking test. The work also includes an analysis of the oral bioavailability and toxicity of the compounds studied.

Experimental

In silico screening of secondary metabolites from species of the genus Spondias

Selecting the molecule database

To select the bank of molecules, a bibliographic survey was carried out using the Science Direct and PubMed databases as search tools. The terms searched were “Spondias and isolations” or “Spondias and isolated”, with the aim of building a database of secondary metabolites that had already been identified and/or isolated from fixed extracts of species of the genus Spondias. This made it possible to specify the species, the compounds and the part of the plant where the compound was identified and/or isolated. Subsequently, the rational design of the 104 molecules was carried out using the ACD/ChemSketch software (version 2016.2.2),²¹ which made it possible to observe the structure of each of the molecules in twoand three-dimensional formats, in order to better understand the molecular structure of the compounds.

In silico generation of molecular descriptors

The potentials of the compounds obtained from the database as drug candidates were analyzed using Lipinskin’s Rule of Five, using the computer program Molinspiration, available online.²² Using this tool, it is possible to calculate the values of physicochemical properties such as the partition coefficient (milogP), topological polar surface area (TPSA), number of hydrogen bond acceptors (nON) or hydrogen bond donors (nOHNH) and molecular mass (MM) associated with Lipinski’s Rule of Five and solubility. Solubility (log S) was calculated using the Virtual Computational Chemistry Laboratory program.²³

Preparing the ligands

Following the methodology proposed by Okada Júnior et al.,²⁴ one of the first steps in a virtual screening is to check the protonation state of the molecule and the pH at which the target (receptor) works. A literature review was therefore carried out on the pH at which MAO-A works. From this, it was identified through the literature review that MAO-A works at physiological pH (7.4), which is generally used for enzymes/proteins - except in some cases, where enzymes/proteins work in more acidic or more alkaline media. After determining the pH range in which MAO-A acts, the molecular design program Marvin Sketch 5.0.9²⁵ was used to check the protonation state of the molecules (ligands).

Next, the freeware software Avogrado® (version:1.2.0, library version:1.2.0, Open Babel version:2.3.90, Qt version: 4.8.6)²⁶ was used, configured to perform classical force field calculations Merck Molecular Force Field 94 (MMFF94), in order to obtain the lowest energy conformation.

Molecular docking protocol

The molecular docking calculations between the ligands and the target (receptor) were carried out using the software Gold Suite 5.1 (Genetic Optimization for Ligand Docking 5.1) (CCDC Software Limited).²⁷ For this study, the crystallographic structure of MAO-A (PDB: 2BXR) was used, having previously been taken from the Protein Data Bank (PDB).²⁸ The chosen crystallographic structure (PDB: 2BXR) was selected due to its high resolution and representation of the MAO-A enzyme in an active state, providing complete and reliable data for the docking analysis. In addition, the optimized structures of the ligands were also used.

MAO-A was complexed with flavin adenine dinucleotide (FAD) (X = 33,013472, Y = -29,611472, Z = 32,760528) and was a dimer (domain A and domain B). In this way, MAO-A was treated by removing crystallized inhibitors, water molecules and other coordinated molecules. In addition, hydrogen atoms were added to the structure of the MAO-A enzyme for the docking simulations.

Analysis of results

The GOLD (Genet Optimization for Ligand Docking) software, which works as a genetic algorithm, performs docking calculations, allowing for the flexibility of ligands and works with the optimization of a predefined fitness function - the GoldSCore - which comprises four components: the hydrogen bond energy of the receptor-ligand complex, the van der Waals bond energy, the intramolecular hydrogen bond energy of the ligand and the internal van der Waals energy of the ligand.²⁹,³⁰ The components that make up the GoldScore are expressed in equation 1.

(1) $$\text{GoldScore}=\text{Shb\_ext}+1.375\text{Svdw\_ext}+\text{Shb\_int}+\text{Svdw\_int}$$

According to the equation 1, Shb_ext is the energy score of the protein-ligand hydrogen bonds, Svdw_ext is the score of the protein-ligand van der Waals energies, Shb_int is the contribution to the score due to the intramolecular hydrogen bonds of the ligand, and Svdw_int is the contribution due to the intramolecular tension of the ligand. However, it should be noted that the term Svdw_ext is expressed as a form of empirical correction to enable/facilitate hydrophobic contacts between protein and ligand, and for this reason it is multiplied by the factor 1.375 when the total score is computed.²⁹,³⁰

Thus, the program performs an internal selection of the docking results based on the fitness function and the chosen score (GoldScore). The docking calculations were designed to obtain ten outputs at each stage of the calculation. In addition, the Discovery Studio Visualizer software³¹ was used to analyze the predicted interactions with the amino acids in the active site of the protein-ligand complex.

Validation of docking protocols by re-docking

In a virtual molecular docking screening, the program generates possible positions for the ligand - called poses - and evaluates the energy of each position using a scoring function. After the poses were generated, they were compared with the crystallographic position of the ligand that was crystallized next to its protein and was removed for further docking. This comparison generates a root mean square deviation (RMSD), with an expected value < 2 Å. Values below 2 Å indicate that the ligand was able to reproduce the expected pose or that it came close to it using the protocol employed.³²,³³

Thus, in the validation by re-docking, the molecular docking process was carried out with the ligand itself from the crystal structure of a ligand-protein complex, in order to try to reproduce the original binding model, using the root mean square deviation (RMSD) values that were consulted in GOLD.

Toxicity of drug candidates

In addition to the stages of selection, preparation and analysis of potential ligands for interaction with the MAO-A enzyme, evaluation of the toxicity of the compounds is essential to ensure the safety and efficacy of the potential drugs. The OSIRIS Property Explorer platform was used for this purpose.³⁴

OSIRIS Property Explorer allows the toxicological effect of drug candidates to be determined, providing information on possible toxicity risks, including mutagenic, carcinogenic, tumorogenic, irritant and reproductive effects of compounds. It is important to note that all the platforms used in this research have been duly validated and are widely recognized in the scientific community, as cited by Rodrigues et al.³⁵

In vitro test of ethanol extracts from species of the genus Spondias

Plant material

The species Spondias tuberosa and Spondias purpurea were collected at the Merejo site (7°47’9”S and 38°1’18” W) and the species Spondias mombin at the Quixaba site (7°47’58”S and 38°1’15”W), both located in the district of Jericó in the city of Flores-PE. The collections were carried out in August 2024. The duly determined exsiccates are deposited in the Brazilian Semiarid Herbarium of the Serra Talhada Academic Unit (UFRPE), where they were identified and registered under the numbers HESBRA 6336, HESBRA 6337 and HESBRA 6335, respectively.

Preparation of ethanolic extracts

The leaves of the three species mentioned above were dried in an oven at 40 ºC, then crushed and the powder formed was exhaustively extracted with ethanol at room temperature. The extracts obtained were filtered and evaporated under reduced pressure in a rotary evaporator at 40-45 °C.

Monoamine oxidase-A (MAO-A) activity assay

The effects of the extracts on MAO-A activity in vitro were evaluated using rat brain tissue. For this, adult Wistar rats (60 days old), kept under standard conditions, were obtained from a breeding colony at the Federal University of Pelotas (UFPel), and their use was approved by the UFPel Animal Ethics Committee (project number: 042314/2023-32).

Preparations enriched with brain mitochondria were used according to Soto-Otero et al.³⁶ Briefly, the isolation medium was homogenized at a 1:4 (m/v) ratio with homogenization buffer (0.0168 M Na₂HPO₄, 0.0106 M KH₂PO₄, 0.32 M sucrose, pH 7.4). The homogenate was then centrifuged at 900 ×g for 5 min at 4 °C. The supernatant was centrifuged again at 12,500 ×g for 15 min at 4 °C. The resulting pellet was resuspended in homogenization buffer and centrifuged again at 12,500 ×g for 15 min at 4 °C. The final mitochondrial pellet was reconstituted in a buffer solution containing 0.0168 M Na₂HPO₄, 0.0106 M KH₂PO₄, and 0.0036 M KCl, pH 7.4.

MAO activity was measured using 100 µL tissue samples prepared according to the method described by Krajl³⁷ with modifications by Soto Otero et al.³⁶ To specifically determine the activity of the MAO-A isoform, the tissue samples were incubated at 37 °C for 5 min in the presence of 250 nM pargyline (a selective MAO-B inhibitor). The extracts were dissolved in dimethyl sulfoxide (DMSO). Then, tissue samples received extract concentrations ranging from 50 to 1,000 µg mL^-1, while the vehicle group received DMSO. Clorgyline (250 nM, a selective MAO-A inhibitor) was used as a positive control. After 10 min, the reaction was initiated by adding kynuramine (a substrate) at a final concentration of 90 μM and incubated for 30 min at 37 °C. The reaction was stopped by adding 10% trichloroacetic acid, and the samples were centrifuged at 16,000 × g for 5 min. To determine MAO activity, 2 mL of 1 M NaOH was added to 800 μL of the supernatant.

The reaction product, 4-hydroxyquinoline (4-HQ), was measured using a fluorimeter at an excitation wavelength of 315 nm and an emission wavelength of 380 nm. Results were expressed as nmol of 4-HQ per milligram of protein per min. For each extract, the result was obtained from three independent experiments (n = 3). Protein content was determined using the Bradford method³⁸ with albumin as the standard.

Statistical analysis

Statistical analyses were performed using the GraphPad Prism software (version 8.2.0.)³⁹ and the results are presented as mean ± standard error of the mean (SEM). D’Agostino-Pearson normality test was used to verify if the data were normally distributed. The data were analyzed by one-way analysis of variance (ANOVA) with the Newman Keuls post hoc test. Values of p < 0.05 were considered statistically significant.

Analysis of the chemical profile of ethanolic extracts of Spondias species by liquid chromatography coupled with mass spectrometry (LC-MS/MS)

The metabolomic analyses of the ethanolic extracts of the Spondias mombin, Spondias tuberosa and Spondias purpurea species were carried out at the Multiuser Characterization and Analysis Laboratory of the Federal University of Paraíba.

15 microlitres of each extract (200 µg mL^-1) were injected into the LC-40D X3 (Shimadzu) equipped with the CBM-40, DGU-40S, LC 40D X3, SIL-40C X3 and CTD-40S modules, coupled to an LCMS-9050 mass spectrometer (Shimadzu) with an electrospray ionization (ESI) source and a quadrupole time-of-flight (Q-TOF) analyzer. LC experiments were carried out using a C18 column (Kromasil-250 mm × 4.6 mm × 5.0 µm). The mobile phase consisted of a 0.1% v/v formic acid/water solution (solvent A) and methanol (solvent B). A linear gradient (5-100% B) was used for elution in 60 min, with a flow rate of 0.6 mL min^-1 and a temperature of 40 ºC. The analysis parameters for mass spectrometers were as follows: capillary voltage: -3.0 kV, nebulizer gas flow of 3.0 L min^-1, drying gas with a flow of 10 mL min^-1 and interface temperature of 300 ºC. The substances were analyzed in negative ionization mode (m/z 100-1200). The MS/MS data were obtained by data-dependent acquisition (DIA) with a collision energy ramp of 5-55 V. TOF-MS observed product ions from 50-1200 m/z.

Results and Discussion

MAO-A is found complexed with FAD as a dimer, the cofactors that control the functioning of the enzyme, and are named A FAD600 and B FAD600. Docking with the cofactor (B FAD600) for the database of compounds identified and/or isolated from species of the genus Spondias showed good results. Notably, the compounds evaluated consistently exhibited scores higher than 56.68, suggesting a high-quality interaction between the binding molecules and the target protein. This positive trend resembles the efficacy observed with fluoxetine, a widely used antidepressant whose therapeutic effects are attributed, in part, to its strong interaction with the relevant molecular targets.

In addition, we performed docking of fluoxetine with MAO-A, obtaining a result of 51.15, demonstrating an effective similarity with the drug already used for depression. This similarity in efficacy between the compounds tested as MAO-A inhibitors and fluoxetine reinforces the therapeutic potential of these compounds as antidepressant agents.

Fluoxetine was chosen as a reference compound instead of the traditional MAO-A inhibitors because it is a drug that is widely used as an antidepressant and has already demonstrated good clinical results. By analyzing its functionality in relation to MAO-A, it is possible to compare its performance with the compounds selected in this study. In subsequent analyses, it will be evident that fluoxetine has molecular descriptor values that are equivalent to those of the compounds studied. This makes it possible to evaluate the potential efficacy of new drug candidates in comparison to an established and successful drug, providing a solid basis for a better understanding of the pharmacokinetic and pharmacodynamic properties of the compounds studied and their potential therapeutic applications.

Furthermore, in order to prove the efficiency of MAO inhibitors derived from natural products, which are selective and reversible, these are considered safer alternatives compared to traditional MAO inhibitors (MAOIs).⁴⁰ Molecular docking was carried out using chlorobemide, the first reversible and selective inhibitor of MAO-A, obtaining a result of 52.30. This value was compared to the studied compounds, which presented a higher score of 56.68, demonstrating the high-quality interaction between the ligand molecules and the target protein.

This result indicates that the higher the fitness value, the greater the interaction between the target and the ligand, forming a more stable complex, which indicates that the ligand molecule is more effective at fitting into the active site of the enzyme. In other words, the energy of the score function interaction calculation makes it possible to choose the binding poses that are theoretically closest to the “real” binding pose.¹⁹ The ligands that obtained the best scores in the formation of complexes with MAO-A are shown in Table 1. All the compounds found in the Spondias genus, including those not selected for Table 1, are presented in detail in the Supplementary Information (SI) section.

Table 1
Molecular docking results of the selected compounds with MAO-A

Compound	Score
β-Carotene-15,15’-epoxide	98.08
Rutin	97.17
Trigaloyl glucose	96.80
Chlorogenic acid butyl ester	93.43
Zeaxanthin	92.10
Phytofluene	84.32
Astaxanthin	80.09
β-Carotene	76.76
Chlorogenic acid	75.70
7,7’,8,8’-Tetrahydro-beta, beta-carotene	74.59

In order to ensure greater precision in the docking, the re-rank score function was inserted, which made it possible to identify the most promising docking solutions, based on the algorithms established by the docking.⁴¹ According to the re-rank score, the three best docking poses were for beta-carotene-15,15’-epoxide, rutin and trigaloyl glucose, respectively; the complexes with the best docking poses belong to the species Spondias mombin L., Spondias tuberosa and Spondias purpurea, respectively.

According to Arrúa,⁴² a score function can be considered efficient when it works well in three applications (binding mode, binding affinity and virtual screening). Thus, the compounds described in this study obtained good score results, since the molecular docking method used works by optimizing the predefined fitness function - the GoldScore, which has functionality in all three applications, allowing a complete analysis of protein-ligand interactions in order to find the best interactions between the target and the ligand, as well as forming a stable complex.¹⁹

As already mentioned, the software used to carry out the docking process made it possible to insert new functions, such as the re-rank score, which made it possible to identify the most promising docking solutions using the algorithms established by the docking process. In this way, instead of using two different algorithms, as done by Brito,⁴³ a single piece of software was able to accurately predict the scores of the compounds that obtained the greatest protein-ligand interaction, considering the degrees of freedom of the side groups of the amino acid residues in the active site during the docking procedure.⁴¹-⁴⁴

In addition, previous studies have investigated rutin and chlorogenic acid in relation to MAO. Rutin showed significant inhibitory activity against MAO-A in in vitro tests, as reported by Engelbrecht et al.⁴⁵ Chlorogenic acid has been studied for its neuroprotective properties, including MAO inhibition, also through in vitro tests, as shown by Grzelczyk et al.⁴⁶ These studies corroborate our findings, reinforcing the relevance and potential of these compounds as MAO-A inhibitors.

Analysis of interactions with amino acids in the active site of the protein-ligand complex

The interactions that occur between a protein and a given ligand result in the formation of a protein-ligand complex, if the ligand has binding affinity and specificity with the protein. Thus, the presence of intermolecular interactions such as hydrophobic, π-π, van der Waals forces, hydrogen bonds, ionic or electrostatic interactions are responsible for indicating the affinity and specificity between the receptor, in this case MAO-A, and the ligands.⁴⁷

After searching the literature and using the Discovery Studio Visualizer software, it was possible to see that the crystallographic structure of MAO-A has 20 amino acid residues that are present in the active site of this enzyme (Tyr-69, Gln-74, Val-91, Val-93, Leu-97, Ile-180, Asn-181, Ile-207, Phe-208, Ser-209, Val-210, Glu-216, Cys-323, Ile-325, Ile-335, Leu-337, Met-350, Phe-352, Tyr-407 and Tyr-444).

In this context, of the 104 compounds that were considered possible MAO-A inhibitors, 78 showed interactions with the amino acid residues of the active site of the enzyme. Therefore, the solutions with the highest scores for each complex were selected using the Gold software (Table 1) and the interactions with the amino acids in the protein-ligand complex were then analyzed using the Discovery Studio Visualizer software in order to analyze the affinity and specificity of the complexes. After carrying out the interaction analyses, the compounds were organized into groups according to the similarities in the interactions.

Group one results

This group includes the compounds which had a similar number of amino acids interacting with amino acids in the enzyme’s active site and also had the highest number of antagonistic interactions, π-π-T-shaped and the presence of the only complex which has a π-cation interaction which, among some of its functions, enzyme catalysis can be highlighted. This means that this ligand can enable the enzyme to act as a biological catalyst in order to increase the speed of reactions in the body. This could be a positive factor, since it could promote a faster increase in the concentration of neurotransmitters in the synaptic cleft, and consequently increase the level of serotonin and noradrenaline, triggering faster activity as an antidepressant. However, if the MAO-A enzyme works more efficiently, the concentrations of neurotransmitters in the synaptic cleft should be lower, as MAO-A degrades neurotransmitters such as serotonin, norepinephrine and dopamine. Inhibiting MAO-A increases the concentrations of these neurotransmitters, which can have a positive therapeutic effect in the treatment of conditions such as depression.

Thus, the analysis showed that the complex with the beta-carotene-15,15’-epoxide ligand (Figure 1) had the highest score and interacted with eight amino acids (Leu337, Ile180, Phe208, Ile335, Phe352, Tyr69, Tyr444, Tyr407). Of these interactions that occurred, six are π-alkyl interactions (Phe208, Leu337, Ile180, Phe352, Tyr69, Ile335) and two are of the alkyl type (Tyr444, Tyr407). These interactions enable the formation of hydrophobic bonds, since in π-alkyl bonds there is an interaction between the electron cloud on an aromatic group and any electron group on the alkyl group. In addition, alkyl type interactions also occur between alkyl groups and consequently also generate hydrophobic interactions.⁴⁸

Figure 1
MAO-A interacting with beta-carotene-15,15’-epoxide.

The interaction with rutin (Figure 2), obtained eight interactions with the amino acids (Tyr69, Try407, Tyr444, Glu216, Ile207, Phe208, Phe352, Gln74), four of which were van der Waals interactions (Tyr444, Tyr407, Glu216, Ile217). In particular, Tyr444 also has a π-π-T-shaped interaction, which is an interaction between two aromatic rings, i.e., the π-π-T-shaped interaction occurs with the interaction of the lateral electron cloud of one ring and the electron cloud of the other ring. Furthermore, this interaction occurs in a T-shape.⁴⁸ It also showed a π-alkyl interaction with the amino acid (Phe208) and three antagonistic interactions with the amino acids (Tyr69, Gln74, Phe352) which indicate repulsion between the amino acid residues with bonds that are both interacting and moving away from the ligand and the active site.⁴⁹

Figure 2
MAO-A interacting with rutin.

Figure 3, in turn, shows the interactions of trigaloyl glucose, in which it is possible to see the presence of a π-cation interaction which is of paramount importance in molecular recognition and enzymatic catalysis.⁴⁸ It also shows interactions with amino acids (Ile180, Asn181, Try407) that participate in the formation of different chemical bonds, including conventional hydrogen bonds and non-conventional hydrogen bonds, as well as the presence of an antagonistic interaction with the amino acid Tyr444.

Figure 3
MAO-A interacting with trigaloyl glucose.

With regard to the hydrogen bonding interactions related to the docking between the inhibitors and MAO-A, it was found that of the 78 complexes that showed interactions with the amino acid residues of the enzyme’s active site, 30 complexes showed hydrogen bonding interactions, which are important interactions that occur in biological systems and are responsible for maintaining the protein’s structure.⁵⁰ In addition, 48 complexes showed π-alkyl, alkyl, π-sulfur and π-π-T-shaped interactions.

However, the presence or absence of interactions of amino acids by hydrogen bonds with the protein-ligand complexes is due to the fact that these interactions are only possible, for the most part, if the inhibitors have hydroxyl groups, or, in certain cases, they also occur due to the presence of nitrogen and fluorine molecules, as they are present in the compounds as MAO-A inhibitors, which interact with the hydrogen atoms of the amino acids.

On the other hand, when interactions occur at the atomic level between the ligand and the protein, it is the electrons that are present in the formation of non-covalent or covalent bonds. Thus, the π-alkyl, alkyl and π-sulfur interactions present in the complexes with MAO-A amino acids belong to the broad category of non-covalent interactions. As already mentioned, π-alkyl interactions have the interaction of the electron cloud and an electron group of any alkyl group on an aromatic group. In the case of π-sulfur interactions, there is a π electron cloud from the aromatic ring that interacts with the two lone pairs from the electron cloud of the sulfur atom and the alkyl-type interactions, in turn, occur as already mentioned between alkyl groups.⁴⁸ Among the complexes presented, it can be seen that alkyl and π-alkyl interactions prevail.

Group two results

This group includes compounds with mostly π-alkyl and alkyl bonds, which favor the formation of hydrophobic bonds due to the presence of non-polar chains or subunits that are solvated by layers of water molecules. Thus, the proximity of these hydrophobic surfaces promotes the collapse of the organized structure of the water molecules, favoring the ligand-receptor interaction at the cost of the entropic gain associated with the disorganization of the system.⁵¹,⁵²

In the case of chlorogenic acid butyl ester, Figure 4, a conventional hydrogen bond with the amino acid (Tyr69), an alkyl interaction with the amino acid (Phe352) and another of the π-alkyl type with the amino acid are observed (Tyr407).

Figure 4
MAO-A interacting with chlorogenic acid butyl ester.

Zeaxanthin, Figure 5, showed hydrogen-carbon interactions (Val210), π-alkyl and alkyl bonds, both of which carried out the same interactions with eight amino acids (Tyr407, Phe208, Tyr69, Phe352, Tyr444, Ile335, Leu337 and Cys323).

Figure 5
MAO-A interacting with zeaxanthin.

Phytofluene, Figure 6, interacted with nine amino acids (Ile325, Leu337, Cys323, Ile335, Phe208, Tyr69, Phe352, Tyr407, Tyr444), with π-alkyl and alkyl interactions, as well as an antagonistic interaction with the amino acid Leu97.

Figure 6
MAO-A interacting with phytofluene.

Astaxanthin, Figure 7, obtained a carbon-hydrogen bond interaction (Ala272). Although this type of C-H interaction is weak (around 4 kJ mol^-1), it is likely that many CH groups participate in the interaction with the π system, increasing the total energy, since this type of interaction persists even in polar media such as water, so it is clear that these interactions can play an important role in the structural stability of organic macromolecules, i.e., they can enable stability in the interactions between the ligand and receptor.⁵³ Astaxanthin also has four alkyl interactions with amino acids (Tyr407, Tyr444, Phe352, Phe208) and three of the π-alkyl type (Ile335, Cys323, Tyr69).

Figure 7
MAO-A interacting with astaxanthin.

Figure 8 shows that β-carotene had six alkyl interactions (Leu337, Ile335, Ile325, Tyr69, Tyr444, Tyr407) and two π-alkyl interactions (Cys323, Phe208).

Figure 8
MAO-A interacting with β-carotene.

Chlorogenic acid, Figure 9, interacted with eight amino acids (Tyr407, Ile335, Cys323, Phe208, Leu337, Tyr69, Phe352, Try444), four of which were alkyl interactions (Leu337, Phe208, Tyr407, Ile335) and four π-alkyl interactions (Tyr444, Ile335, Cys323, Tyr69). However, the affinity of the alkyl bonds with carbon-hydrogen bonds was found to exist, since a detailed analysis of the coupling of these bonds revealed that the amino acids (Phe208, Leu337, Tyr407, Ile335) showed a degree of influence on the affinity energy of the interactions.

Figure 9
MAO-A interacting with chlorogenic acid.

Finally, the 7,7’,8,8’-tetrahydro-beta, beta-carotene complex, Figure 10, showed five alkyl interactions with the amino acids (Tyr69, Phe352, Ile335, Ile325, Leu337), two π-alkyl interactions with the amino acids (Cys323, Tyr407, Leu97) and one carbon-hydrogen bond interaction (Val210).

Figure 10
MAO-A interacting with 7,7’,8,8’-tetrahydro-beta, beta-carotene.

Validation of the re-docking docking protocol

The re docking docking protocol was validated using MAO-A’s own crystallographic ligand (PDB ID: 2BXR), in order to confirm the reliability of the docking parameters used in this study. In this procedure, the ligand originally co-crystallized with the enzyme was removed, and the docking process was repeated to verify the ability of the method to reproduce the known binding pose (Figure 11).

Figure 11
Overlay of the monoamine oxidase A (MAO-A) ligand and the best docking pose with RMSD equal to 0.3 Å.

The RMSD value obtained was 0.3 Å, indicating that the predicted binding pose is very close to the original crystallographic position, demonstrating high accuracy and reliability of the docking protocol. In general, an RMSD value equal to or less than 2 ų³ is considered acceptable for successful validation of re docking, as it suggests that the method is capable of reproducing the experimentally observed binding mode.

This validation step is critical because it ensures that the docking parameters, including scoring functions and search algorithms, are properly calibrated to predict accurate protein-ligand interactions. In this way, the docking results presented in this study can be considered reliable and robust for the identification of potential MAO-A inhibitors from the Spondias genus.

Analysis of molecular descriptors

After analyzing the molecular docking results, it is crucial to consider the role of molecular descriptors in the selection and characterization of the compounds studied. Molecular descriptors are quantitative parameters that represent various structural and physicochemical properties of molecules, providing essential information for understanding their biological activity.⁵⁴

The molecular descriptors calculated included partition coefficient (log P), topological polar surface area (TPSA), number of hydrogen bond donors and acceptors, molecular mass and solubility. These physicochemical properties are crucial for assessing the compounds’ ability to interact with MAO-A and influence its inhibitory activity.

In addition, the molecular descriptors were essential for obtaining the lowest energy conformation of the molecules, ensuring that they were in a suitable form to interact with the active site of the target enzyme. This analysis contributed significantly to the selection of the most promising drug candidates, integrating structural and physicochemical information with the molecular docking results. The compounds shown in Table 2 were the ones that showed the best results in molecular docking, highlighting their relevance for future drug development studies.

Table 2
Molecular descriptors of the compounds selected with the best results in molecular docking with the MAO-A enzyme and fluoxetine as a reference

Compound	MW / (g mol-1)	log P	nON	nOHNH	Violations	log S	TPSA / Å2	BBB
Beta-carotene-15,15’-epoxide	552.89	9.73	1	0	2	-6.09	12.53	no
Rutin	610.52	-1.06	16	10	4	-	269.43	no
Trigaloyl glucose	636.47	0.64	18	11	4	-2.74	310.66	no
Chlorogenic acid butyl ester	426.46	2.30	9	5	1	-2.87	153.75	no
Zeaxanthin	568.89	9.39	2	2	2	-5.94	40.46	no
Phytofluene	542.94	10.08	0	0	2	-6.19	0.00	no
Astaxanthin	596.85	8.60	4	2	2	-5.95	74.60	no
β-Carotene	536.89	9.84	0	0	1	-	0.00	no
Chlorogenic acid	324.33	0.68	7	5	0	-2.24	127.44	no
7,7’,8,8’-Tetrahydro-beta, beta-carotene	572.92	9.40	2	2	2	-6.02	40.46	no
Fluoxetine	309.33	4.53	2	1	0	-5.26	21.26	yes

After analyzing the results of the molecular docking and the molecular descriptors, another crucial aspect to consider is the ability of the compounds to be absorbed and effectively penetrate biological systems. A parameter widely used to assess this capacity is the partition coefficient (log P), which expresses the sensitivity of the compound to lipids and its tendency to pass through biological membranes. According to the literature, log P values below 5 indicate good absorption, which is essential for the efficacy of a potential drug.⁵⁵

After analyzing the log P, another relevant parameter to be considered is the polar area (PSA), which is related to the assessment of cell permeability and in vivo bioavailability of an agent.⁵⁶ Thus, while 75.96% of the compounds selected have an adequate log P, suggesting a good absorption capacity, 52% of them also showed good cell permeability, according to the PSA analysis. These compounds include pheophorbid-a, quercetin, epicatechin, chlorogenic acid and epigallocatechin. However, only rutin, trigaloyl glucose and chlorogenic acid butyl ester, which were subjected to molecular docking, meet the criteria for adequate log P, with values below 5, and PSA values within the permitted limit of 140 Å.

In previous works,⁵⁷,⁵⁸ the PSA value is of great importance with regard to the antibacterial action of a compound because it establishes that the higher the PSA value, the greater the effect against Gram-positive bacteria. In this sense, these same authors report that a higher molecular mass, within the range as previously reported,⁵⁹ is associated with greater polarity (lower log P value) and greater activity against Gram-negative bacteria. Thus, it was observed that 65.38% of the compounds analyzed in this study had a molar mass (MM < 500) within the limits established by the Rule of Five.

Considering these parameters, analyzing the molar mass of the compounds is also relevant to assessing their suitability as drug candidates. Of the compounds listed in Table 2, chlorogenic acid butyl ester and chlorogenic acid meet this molar mass criterion (MM < 500). However, compounds 7,7’,8,8’-tetrahydro-beta, rutin, trigaloyl glucose and astaxanthin have a molar mass slightly higher than the established limit, indicating the need to consider factors other than molecular mass when assessing their suitability as drug candidates.

As a result of observing the values for the number of hydrogen bond donors (nOHNH ≤ 5) and number of hydrogen bond acceptors (nON ≤ 10), 53, 85% and 65, 38% of the compounds were found, respectively. This was established using the rule from de Medeiros Filho⁴⁹ and is within the limits. This rule also establishes that two or more violations for a compound can represent bioavailability problems.⁶⁰

Among the compounds listed in Table 2, only a few meet the criteria established for the number of hydrogen bond donors (nOHNH ≤ 5) and the number of hydrogen bond acceptors (nON ≤ 10). Beta-carotene-15,15’-epoxide has only one hydrogen bond donor (nOHNH = 1) and no hydrogen bond acceptors (nON = 0), meeting both criteria. Chlorogenic acid butyl ester, on the other hand, has five hydrogen bond donors (nOHNH = 5) and nine hydrogen bond acceptors (nON = 9), also meeting the established criteria. Zeaxanthin has two hydrogen bond donors (nOHNH = 2) and two hydrogen bond acceptors (nON = 2), also within the established limits.

In addition, the solubility of organic compounds is usually represented by log S, where S is the concentration of the compound in M. In practice, around 85% of drugs have log S values between -1 and -5 and virtually no values below -6.⁶¹,⁶² Values above -1 are associated with very polar molecules such as sugars or small peptides, have low membrane permeability in the absence of active transport, but are not problematic.

Of the compounds analyzed, 57, 69% showed values between -1 and -6. Empirically, it is apparent that the log S target between -1 and -5 for most drugs reflects the relationship between the polarity required for reasonable aqueous solubility and the hydrophobicity required for acceptable membrane passage.⁶¹ Of the compounds listed, all meet the criterion, as they all have log S values within the typical solubility range for drugs, which is between -1 and -5, with the exception of phytofluene and beta-carotene-15,15’-epoxide, which have values below -6.

An additional important parameter is the blood brain barrier (BBB) penetration index. This parameter indicates the ability of compounds to cross the blood-brain barrier, which is crucial for drugs intended to treat neurological diseases. The BBB is a highly selective barrier that protects the central nervous system (CNS) from potentially harmful substances present in the bloodstream. For a compound to be effective in treating neurological conditions, it must be able to cross this barrier.⁶³ TPSA values below 60-70 Ų are often associated with greater BBB permeability.⁶⁴ In Table 2, all listed compounds do not have the ability to penetrate the BBB, except fluoxetine, used as reference, which is known for its ability to cross this barrier. The lack of penetration into the BBB is important, as these compounds may be safer and less likely to cause side effects in the CNS, when not used for non-neurological treatments, which would not be the case in our study, although these results were not promising, more trials should be carried out with these compounds.

These results reinforce the importance of the compounds selected as potential drug candidates, highlighting their structural and physicochemical diversity and providing a solid basis for evaluating their therapeutic efficacy.

Toxicity analysis of the best performing compounds

After selecting the best performing ligands in the docking simulations, a comprehensive evaluation of the molecular descriptors of these compounds was carried out. Then, using the OSIRIS Property Explorer platform, a detailed analysis of their toxicity was carried out. This tool made it possible to determine the potential toxicological risk of the ligands, including mutagenic, carcinogenic, tumorogenic, irritant and reproductive system effects.⁶⁵

The compounds were carefully analyzed, considering both their efficacy as potential drugs and their safety in terms of toxicity. The results indicated that the selected ligands have a favorable safety profile, with low potential for adverse effects.⁶⁵

These results have significant implications for the development of new drugs. Ligands that combine high affinity for the MAO-A enzyme with low toxicity potential are considered preferable for moving on to the next stages of research and development. These compounds have the potential to become promising candidates for the treatment of MAO-A-associated diseases.

For a clear and concise visualization of the toxicity profile, the best performing compounds have been organized in Table 3, highlighting their mutagenic, carcinogenic, tumorogenic, irritant and reproductive system effects. In addition, a comparison was made with fluoxetine (Table 3), a drug widely used in the treatment of depression, to assess the relative toxicity profile of the ligands in comparison with a reference compound.

Table 3
Toxicity profile of the compounds with the best docking performance with MAO-A and comparison of toxicity with fluoxetine

Compound	Mutagen	Tumorigenic	Irritating	Reproductive effect
Beta-carotene-15,15’-epoxide	low	low	low	low
Rutin	low	low	low	low
Trigaloyl glucose	medium	medium	low	low
Chlorogenic acid butyl ester	low	low	low	low
Zeaxanthin	low	low	low	low
Phytofluene	low	low	low	low
Astaxanthin	low	low	low	high
β-Carotene	low	low	low	low
Chlorogenic acid	low	low	low	low
7,7’,8,8’-Tetrahydro-beta, beta-carotene	low	low	low	low
Fluoxetine	low	low	low	low

After selecting the ligands with the best performance in docking simulations with MAO-A, we carried out a comprehensive analysis of their toxicity using the OSIRIS Property Explorer platform. The results indicated a favorable safety profile for most of the compounds, with low potential for adverse effects, making them promising candidates for the development of new drugs.

When comparing the compounds with fluoxetine, a drug widely used in the treatment of depression, we observed that most of the ligands showed comparatively low toxicity in all aspects evaluated, including mutagenicity, tumorigenicity, irritation and effects on the reproductive system, similar to fluoxetine.

However, some compounds, such as astaxanthin, stood out for having a relatively high effect on the reproductive system, which may require further investigation of their safety for potential use as a drug.

In summary, the selected ligands exhibited a promising toxicity profile, with most showing comparatively low toxicity compared to fluoxetine, suggesting their potential as candidates for the development of new drugs for the treatment of MAO-A-associated diseases.

Monoamine oxidase inhibition

In vitro MAO-A inhibition tests were carried out with the ethanolic extracts of the Spondias mombin (EEC), Spondias tuberosa (EEU) and Spondias purpurea (EES) species, which were the species that reported the presence of the constituents that performed best in molecular docking. These tests sought to confirm, experimentally, what had been observed about the inhibition of MAO-A computationally.

The effects of the extracts on cerebral MAO-A activity are depicted in Figure 12. A one-way ANOVA revealed significant effects in experiments for the three tested extracts: EES (F_{(7, 16)} = 111.4, p < 0.0001) (Figure 12a), EEC (F_{(7, 16)} = 49.9, p < 0.0001) (Figure 12b), and EEU (F_{(7, 16)} = 54.54, p < 0.0001) (Figure 12c).

Figure 12
Effects of the EES (a), EEC (b), and EEU (c) extracts on cerebral MAO-A activity in vitro. MAO-A activity is expressed as nmol of 4-hydroxyquinoline per milligram of protein per min. Values are reported as the mean ± SEM (n = 3). Clorgyline (250 nM, a selective MAO-A inhibitor) was used as a positive control (C+). One-way ANOVA/Newman-Keuls: *p < 0.05, **p < 0.01, and ***p < 0.001 as compared with the control (C) group, and ^#p < 0.05, ^##p < 0.01 and ^###p < 0.001 as compared with the vehicle (V) group.

Post hoc analyses demonstrated that the vehicle group (DMSO) did not differ from the control group in any of the experiments (p > 0.05). As shown in Figure 12a, the EES extract reduced cerebral MAO-A activity at concentrations of 500 and 1,000 µg mL^-1 compared with the vehicle group (DMSO) (p < 0.05 and p < 0.001, respectively). Additionally, the EEC (Figure 12b) and EEU (Figure 12c) extracts also inhibited MAO-A activity at concentrations of 500 and 1,000 µg mL^-1 (for both p < 0.01 and p < 0.001, respectively). Moreover, the positive control clorgyline (a selective MAO-A inhibitor) caused a significant inhibition of MAO-A activity (p < 0.001), validating the technique.

This inhibition follows a dose-dependent relationship, in which higher concentrations of the inhibitor provide greater efficacy up to a certain saturation point. This was observed in studies with natural compounds such as galangin and apigenin, which showed a more pronounced inhibition of MAO-A and MAO-B at higher doses, due to the greater availability of molecules to block the enzyme’s active site and prevent the degradation of neurotransmitters.⁴⁰,⁶⁶ These mechanisms are widely studied in the context of neurodegenerative diseases and depression, illustrating the pharmacological potential of plant extracts at appropriate concentrations.⁶⁶

Analysis of the chemical profile of the ethanolic extracts of the Spondias species

The characterization of the chemical profile of the ethanolic extracts of the S. mombin, S. purpurea and S. tuberosa species is essential to confirm the presence of the compounds previously identified as potential MAO-A inhibitors in the molecular docking studies. This analysis has made it possible to establish a more precise correlation between the compounds identified in silico and the experimental data obtained in the enzyme inhibition tests, ensuring that the compounds responsible for the biological activity are correctly identified.

The extracts were analyzed by LC-MS/MS. This method enabled the identification of various phytochemical compounds present in the leaves of these species, predominantly phenolic constituents. The structural characterization of these compounds was carried out by interpreting their MS² fragmentation spectra, combined with comparisons with reference compounds from spectral databases and data from scientific literature. The mass spectrometry analysis was carried out in negative ionization mode, which was selected due to its high sensitivity for the detection of phenolic compounds.⁶⁷

The chromatographic profiles obtained revealed the presence of several phenolic compounds, including flavonoids, which were common to the three Spondias species analyzed. Each species showed unique and shared peaks, suggesting similarities in their phytochemical composition. The chromatograms of each species are shown in Figures S1, S7 and S12, and the chemical constituents identified are detailed in Tables S1, S2 and S3 (SI section).

The LC-MS/MS analysis revealed that two of the best docking scores, such as chlorogenic acid, with m/z [M - H]^- 353.1 and a retention time of 19.36 min, were present in both S. purpurea and S. mombin samples. The MS/MS fragmentation of this compound revealed a fragment ion at m/z 191, corresponding to the quinic acid part resulting from the loss of caffeic acid, while the fragment at m/z 85 indicates a further loss of CO from quinic acid.⁶⁸,⁶⁹ However, in S. purpurea, quinic acid (m/z 191) and chlorogenic acid (m/z 353) coeluted under the chromatographic conditions employed, with overlapping retention times, but the MS/MS spectra showed characteristic fragments of both compounds, even though the chromatographic separation was insufficient to completely resolve the corresponding peaks.

The other compound of the best docking score, rutin, was present in all three samples of Spondias species. Rutin, with a retention time of 30.32 min and m/z 609.1, exhibits characteristic fragments in m/z 301 and m/z 151. The m/z 301 fragment corresponds to a desglycosylation suffered due to the loss of m/z 308, and the fragment at m/z 301 suffered a cleavage through Retro-Diels-Alder to give the fragment in m/z 151.⁷⁰ Rutin is a glycosylated flavonol of quercetin found in several medicinal plants and is known for its important biological activities, including neuroprotective, anticancer, antioxidant, and anti-inflammatory effects.⁷¹

In addition to chlorogenic acid and rutin, other secondary metabolites were also identified in the ethanolic extracts of the Spondias species, including gentesic acid 5-O-glucoside (m/z 315.1),⁶⁷,⁶⁸ kaempferol-3-O-β-D-glucoside-7-O-α-L-rhamnoside (m/z 593.1),⁷² and gingerglycolipid A (m/z 675.3).⁷³ Similarly to rutin, these compounds were present in all Spondias samples and were identified through comparison with reference libraries and characteristic fragmentation patterns reported in the literature.

Conclusions

A total of 104 compounds were identified in 21 species of the Spondias genus, with the aim of studying the interactions between protein-ligand complexes, as well as trying to understand the energetic interactions, intermolecular interactions and the affinity between the receptor and the ligand, in order to find new drug candidates with high affinity and specificity for the target site and with greater intrinsic activity.

Thus, the molecular docking study made it possible to observe that the compounds identified and/or isolated from the species of the genus Spondias obtained scores higher than 56.68, which predicts that these secondary metabolites (ligands) made stable interactions with MAO-A (target), forming stable complexes.

In the analysis of interactions with amino acids in the active site of the protein-ligand complex, it was found that of the 104 complexes as possible inhibitors, more than half - a total of 78 complexes - showed interactions with amino acid residues in the active site of MAO-A, with the majority of these compounds belonging to species of the genus Spondias mombin L., Spondias tuberosa and Spondias purperea. Thus, as already mentioned, 48 complexes showed π-alkyl, alkyl, π-sulfur and π-π-T-shaped interactions and 30 complexes showed hydrogen bond interactions, which shows the existence of affinity and specificity of the ligands with monoamine oxidase A (MAO-A).

Among the ten complexes with the best scores: beta-carotene-15,15’-epoxide (98.08), rutin (97.17), trigaloyl glucose (96.80), chlorogenic acid butyl ester (93.43), zeaxanthin (92.10), phytofluene (84.32), astaxanthin (80.09), β-carotene (76.76), chlorogenic acid (75.70) and 7, 7’,8,8’-tetrahydro-beta,beta-carotene (74.59), beta-carotene-15,15’-epoxide (98.08) was considered the most promising of the MAO-A inhibiting compounds, since it had the highest score and made eight interactions with the amino acid residues present in the active site of the enzyme.

However, it should be emphasized that the other compounds also showed promising results as possible MAO-A inhibitors, since they had high scores of more than 56.68 and all the compounds were able to interact with at least five amino acid residues present in the enzyme’s active site, which means high interaction, affinity and specificity between the protein-ligand complexes. Furthermore, the method proved to be effective, as it showed an RMSD of 0.3 Å, which indicates that the ligands were able to reproduce an expected or similar pose with the ligand itself, which was crystallographically bound to the protein.

In addition, the analysis of compliance with Lipinski’s Rule of Five showed that most of the compounds met the criteria for “drug-likeness”, indicating good potential for development as drugs. The criteria include properties such as water solubility and cell permeability, which are crucial for drug efficacy. Toxicity predictions were also carried out, suggesting that the candidate compounds are safe for future stages of development, without showing significant toxicity.

The results of the in vitro tests carried out with the extracts of the species of the genus Spondias showed inhibitory activities on the MAO-A enzyme at concentrations of 500 and 1000 µg mL^-1, resulting in a significant inhibition of the MAO-A enzyme.

Thus, it is clear that the secondary metabolites belonging to the species of the genus Spondias reported in this study make significant and promising contributions to the discovery of new therapeutic agents, which could contribute to the treatment of neurological diseases, including depression. In addition, there is a lack of in vitro studies with species from the genus under study as MAO-A inhibitors in the literature. Therefore, this study can contribute to further studies, where the best inhibitors reported in this study can be tested both in vivo.

Supplementary Information

All supplementary data (mass spectrum, LC-MS/MS data) are available free of charge at http://jbcs.sbq.org.br as PDF file.

Data Availability Statement

All data are available in the text.

Acknowledgments

The authors thank the Federal Rural University of Pernambuco (UFRPE) and the Federal University of Pelotas (UFPel) for their support in executing the project and the Multiuser Characterization and Analysis Laboratory of the Federal University of Paraíba (LMCA-UFPB) for analyzing the chemical profile of the samples.

References

• ¹ Veiga Jr., V. F.; Maciel, M. A. M.; Quim. Nova 2015, 28, 519. [Crossref]
  » Crossref
• ² Pan, S. Y.; Zhou, S. F.; Gao, S. H.; Yu, Z. L.; Zhang, S. F.; Tang, M. K.; Sun, J. N.; Ma, D. L.; Han, Y. F.; Fong, W. F.; Ko, K. M.; Evid. Based Complement. Alternat. Med. 2013, 73, ID 627375. [Crossref]
  » Crossref
• ³ Gobbo-Neto, L.; Lopes, N. P.; Quim. Nova 2007, 30, 374. [Crossref]
  » Crossref
• ⁴ Pavarinia, D. P.; Pavarini, S. P.; Niehues, M.; Lopes, N. P.; Anim. Feed Sci. Technol. 2012, 176, 5. [Crossref]
  » Crossref
• ⁵ Braibante, M. E. F.; da Silva, D.; Braibante, H. T. S.; Pazinato, M. S.; Quim. Nova Escola 2014, 36, 168. [Crossref]
  » Crossref
• ⁶ Fumagali, E.; Gonçalves, R. A. C.; Machado, M. F. P. S.; Vidoti, G. J.; de Oliveira, A. J. B.; Rev. Bras. Farmacogn. 2008, 18, 627. [Crossref]
  » Crossref
• ⁷ Pereira, R. J.; Cardoso, M. G.; J. Biotechnol. Biodiversity 2012, 3, 146. [Crossref]
  » Crossref
• ⁸ Specian, V.; Orlandelli, R. C.; Felber, A. C.; Azevedo, J. L.; Pamphile, J. A.; J. Health Sci. 2014, 16, 345. [Crossref]
  » Crossref
• ⁹ Arroio, A.; Honório, K. M.; da Silva, A. B. F.; Quim. Nova 2010, 33, 694. [Crossref]
  » Crossref
• ¹⁰ Lyne, P. D.; Drug Discovery Today 2002, 7, 1047. [Crossref]
  » Crossref
• ¹¹ Laskowski, R. A.; Chistyakov, V. V.; Thornton, J. M.; Nucleic Acids Res. 2005, 33, 266. [Crossref]
  » Crossref
• ¹² Christenhusz, M. J. M.; Byng, J. W.; Phytotax 2016, 261, 201. [Crossref]
  » Crossref
• ¹³ da Silva, G. A. D.; de Brito, N. J. N.; dos Santos, E. C. G.; Revista BioFarm. 2014, 10, 27. [Link]
  » Link
• ¹⁴ Jia, Z.; Zhu, Q.; Bioorg. Med. Chem Lett. 2010, 20, 222. [Crossref]
  » Crossref
• ¹⁵ Wang, D.; Chen, N.; Taaranto, A. G.; Jing, Y.; Wen, C.; Kong, D. X.; Bioorg. Chem. 2021, 106, 104503. [Crossref]
  » Crossref
• ¹⁶ Korolkovas, A.; Burckhalter, J. H.; Química Farmacêutica, 1^st ed.; Guanabara Koogan: São Paulo, Brazil, 2008.
• ¹⁷ Shih, J. C.; Thompson, R. P.; Am. J. Hum. Genet 1999, 65, 593. [Crossref]
  » Crossref
• ¹⁸ Willner, P.; Krüger, J. S.; Belzung, C.; Neurosci. Biobehav. Rev 2013, 37, 2331. [Crossref]
  » Crossref
• ¹⁹ Verdonk, M. L.; Cole, J.; Hartshorn, M. J.; Proteins Struct. Funct. Bioinforma 2003, 52, 609. [Crossref]
  » Crossref
• ²⁰ Melfi, F.; Carradori, S.; Angeli, A.; D’Agostino, I.; Expert Opin. Drug Discovery 2023, 18, 851. [Crossref]
  » Crossref
• ²¹ ACD/ChemSketch, version 2016.2.2; Advanced Chemistry Development, Inc., Toronto, ON, Canada, 2016.
• ²² Molinspiration Cheminformatics, http://www.molinspiration.com/, accessed in July 2025.
  » http://www.molinspiration.com/
• ²³ Computational Chemistry Laboratory program, http://www.vcclab.org/, accessed in July 2025.
  » http://www.vcclab.org/
• ²⁴ Okada-Júnior, C. W.; Monteiro, G. C.; Aguiar, A. C. C.; Batista, V. S.; de Souza, J. O.; Souza, G. E.; Bueno, R. V.; Oliva, G.; Nascimento-Júnior, N. M.; Guido, R. V. C.; Bolzani, V. S.; ACS Omega 2018, 3, 9424. [Crossref]
  » Crossref
• ²⁵ Marvin Sketch, version 5.0.9; ChemAxon Ltd.: Budapest, Hungary, 2008.
• ²⁶ Avogadro, version:1.2.0, library version:1.2.0, Open Babel version:2.3.90, Qt version: 4.8.6; Kitware, Inc., Clifton Park, New York, USA, 2016.
• ²⁷ GOLD (Genetic Optimization for Ligand Docking), Suite 5.1, CCDC Software Limited, Cambridge, Cambridgeshire, United Kingdom, 2011.
• ²⁸ Colibus, L.; Li, M.; Binda, C.; Lustig, A.; Edmondson, D E.; Mattevi, A.; Proc. Natl. Acad. Sci. 2005, 102, 12684. [Crossref]
  » Crossref
• ²⁹ Jones, G.; Willett, P.; Glen, R. C.; J. Mol. Biol 1995, 245, 43. [Crossref]
  » Crossref
• ³⁰ Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.; J. Mol. Biol 1997, 267, 727. [Crossref]
  » Crossref
• ³¹ Discovery Studio Visualizer, version 4.5; BIOVIA, Dassault Systèmes, San Diego, CA, USA, 2015.
• ³² Pietro-Martínez, F. D.; Arciniega, M.; Medina-Franco, J. L.; TIP, Rev. Espec. Cienc. Quim.-Biol. 2018, 21, 65. [Crossref]
  » Crossref
• ³³ de Oliveira, J. S.; Ramos, L. F. P.; Rev. Neurociênc 2021, 29, 1. [Crossref]
  » Crossref
• ³⁴ OSIRIS Property Explorer platform, http://www.organicchemistry.org/prog/peo, accessed in July 2025.
  » http://www.organicchemistry.org/prog/peo
• ³⁵ Rodrigues, J. S. M.; Costa, E. D.; Res. Soc. Dev. 2021, 10, e53010414459. [Crossref]
  » Crossref
• ³⁶ Soto-Otero, R.; Méndez-Alvarez, E.; Hermida-Ameijeiras, A.; Sánchez-Sellero, I.; Cruz-Landeira, A.; Lamas, M. L.; Life Sci. 2001, 68, 879. [Crossref]
  » Crossref
• ³⁷ Krajl, M.; Biochem. Pharmacol. 1965, 14, 1684. [Crossref]
  » Crossref
• ³⁸ Bradford, M. M.; Anal. Biochem. 1976, 72, 248. [Crossref]
  » Crossref
• ³⁹ GraphPad Prism, version 8.2.0; GraphPad Software Inc., San Diego, CA, USA, 2019.
• ⁴⁰ Chaurasiya, N. D.; Leon, F.; Muhammad, I.; Tekwani, B. L.; Molecules 2022, 27, 4297. [Crossref]
  » Crossref
• ⁴¹ Thomsen, R.; Christensen, M. H. J.; J. Med. Chem. 2006, 49, 3315. [Crossref]
  » Crossref
• ⁴² Arce, O. E. A.: Função de Escore Baseada em Machine Learning para Docagem Molecular Proteína-Ligante; MSc Dissertation, Universidade Federal do Rio Grande, Rio Grande, Brazil, 2020. [Link] accessed in August 2025
  » Link
• ⁴³ Brito, G. S. L.: Corroboração Computacional por “Docking” Molecular da Hipótese Auto-imune da Esquizofrenia; PhD Thesis, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil, 2007. [Link] accessed in August 2025
  » Link
• ⁴⁴ Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.; Belew, R. K.; Olson, A. J.; J. Comput. Chem 1998, 19, 1639. [Crossref]
  » Crossref
• ⁴⁵ Engelbrecht, I.; Petzer, J. P.; Petzer, A.; Cent. Nerv. Syst. Agents Med. Chem 2019, 19, 133. [Crossref]
  » Crossref
• ⁴⁶ Grzelczyk, J.; Budryn, G.; Peña-García, J.; Szwajgier, D.; Gałązka-Czarnecka, I.; Oracz, J.; Pérez-Sánchez, H.; Food Chem. 2021, 38, 129108. [Crossref]
  » Crossref
• ⁴⁷ Guryanov, I.; Fiorucci, S.; Tennikova, T.; Mater Sci. Eng. C 2016, 68, 890. [Crossref]
  » Crossref
• ⁴⁸ Alencar, W. L. M.; Arouche, T. S.; Gomes Neto, A. F.; Ramalho, T. C.; Carvalho Jr., R. N.; Chaves Neto, A. M. J.; Sci. Rep 2022, 12, 3316. [Crossref]
  » Crossref
• ⁴⁹ de Medeiros Filho, F. M.; Nascimento, K. S.; Santos, W. O.; Frazão, N. F.; Educ. Ciênc. Saúde 2020, 7, 1. [Crossref]
  » Crossref
• ⁵⁰ Barreiro, E. J.; Fraga, C. A. M.; Química Medicinal: As Bases Moleculares da Ação dos Fármacos, 3^rd ed.; Artmed: Porto Alegre, Brazil, 2015.
• ⁵¹ Barreiro, E. J.; Fraga, C. A. M.; Química Medicinal: As Bases Moleculares da Ação dos Fármacos, 2^nd ed.; Artmed: Porto Alegre, Brazil, 2008.
• ⁵² Voet, D.; Voet, J. G.; Pratt, C. W.; Fundamentos de Bioquímica: A Vida em Nível Molecular, 4^th ed.; Artmed: Porto Alegre, Brazil, 2014.
• ⁵³ Shanthi, V.; Ramanathan, K.; Sethumadhavan, R.; Appl. Biochem. Biotechnol. 2010, 160, 1473. [Crossref]
  » Crossref
• ⁵⁴ Todeschini, R.; Consonni, V.; Molecular Descriptors for Chemoinformatics, 2^nd ed.; Wiley-VCH: Weinheim, Germany, 2009.
• ⁵⁴ Lesney, M. S.; Today’s Chemist at Work 2004, 5, 26. [Crossref]
  » Crossref
• ⁵⁵ Vieth, M.; Siegel, M. G.; Higgs, R. E.; Watson, I. A.; Robertson, D. H.; Savin, K. A.; Durst, G. L.; Hipskind, P. A.; J. Med. Chem. 2004, 47, 224. [Crossref]
  » Crossref
• ⁵⁶ O´Shea, R.; Moser, H. E.; J Med. Chem. 2008, 51, 2871. [Crossref]
  » Crossref
• ⁵⁷ Mariño, P. A.; Estudo de Chalconas como Antibacterianos Potenciais: Síntese, Avaliação da Ação Antibacteriana e das Propriedades Físico-Químicas; MSc Dissertation, Federal University of Pampa, Bagé, Brazil, 2014. [Link] accessed in August 2025
  » Link
• ⁵⁸ Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.; Adv. Drug Delivery Rev 1997, 23, 3. [Crossref]
  » Crossref
• ⁵⁹ Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.; Adv. Drug Delivery Rev 2001, 46, 3. [Crossref]
  » Crossref
• ⁶⁰ Jorgensen, W. L.; Duffy, E. M.; Adv. Drug Delivery Rev 2002, 54, 355. [Crossref]
  » Crossref
• ⁶¹ Huuskonen, J.; Salo, M.; Taskinen, J.; J. Chem. Inf. Comput. Sci 1999, 38, 450. [Crossref]
  » Crossref
• ⁶² Pardridge, V. M.; J. Cereb. Blood Flow Metab. 2012, 32, 1959. [Crossref]
  » Crossref
• ⁶³ Kelder, J.; Grootenhuis, P. D.; Bayada, D. M.; Delbressine, L. P.; Ploemen, J. P.; Pharm. Res. 1999, 16, 1514. [Crossref]
  » Crossref
• ⁶⁴ Janicka, M.; Sztanke, M.; Sztanke; K.; Molecules 2024, 29, 287. [Crossref]
  » Crossref
• ⁶⁵ Berlonwitz, I.; Egger, K.; Front. Pharmacol. 2022, 2, 886408. [Crossref]
  » Crossref
• ⁶⁶ Chaurasiya, N. D.; Ibrahim, M. A.; Muhammad, I.; Walker, L. A.; Tekwani, B. L.; Molecules 2014, 19, 18936. [Crossref]
  » Crossref
• ⁶⁷ Maggi, F.; Lucarini, D.; Papa, F.; Peron, G.; Dall’Acqua, S.; Biochem. Syst. Ecol. 2016, 66, 50. [Crossref]
  » Crossref
• ⁶⁸ Mngoma, M. F.; Magwaza, L. S.; Mditshwa, A.; Tesfay, S. Z.; Mkhwanazi, B. N.; Nkomo, M. A.; S. Afr. J. Bot. 2025, 181, 162. [Crossref]
  » Crossref
• ⁶⁹ Oliveira, D. S.; Lima, L. S.; Antonio, A. S.; Wiedemanna, L. S. M.; Veiga-Junior, V. F.; Quim. Nova 2020, 43, 72. [Crossref]
  » Crossref
• ⁷⁰ Yang, J.; Qian, D.; Jiang, S.; Shang, E. X.; Guo, J.; Duan, J. A.; J. Chromatogr. B 2012, 898, 95. [Crossref]
  » Crossref
• ⁷¹ Chua, L. S.; J. Ethnopharmacol. 2013, 150, 805. [Crossref]
  » Crossref
• ⁷² MassBank, https://massbank.eu/MassBank/RecordDisplay?id=MSBNK-RIKEN-PR100945, accessed in June 2025.
  » https://massbank.eu/MassBank/RecordDisplay?id=MSBNK-RIKEN-PR100945
• ⁷³ Li, Y.; Zhou, B.; Teng, Z.; Zhang, M.; Yu, L.; Luo, Y.; Chen, P.; Sun, J.; J. Agric. Food Res. 2023, 14, 100719. [Crossref]
  » Crossref

Edited by

• Editor handled this article:
  Paula Homem-de-Mello (Executive)

Publication Dates

• Publication in this collection
  29 Sept 2025
• Date of issue
  2025

History

• Received
  13 June 2025
• Accepted
  29 Aug 2025