Chemical Composition, Antimicrobial Properties of Siparuna guianensis Essential Oil and a Molecular Docking and Dynamics Molecular Study of its Major Chemical Constituent

The essential oil of Siparuna guianensis was obtained by hydrodistillation. The identification of the chemical compounds was performed by gas chromatography coupled with mass spectrometry (GC/MS). Antimicrobial activity was investigated for four microorganisms: Streptococcus mutans (ATCC 3440), Enterococcus faecalis (ATCC 4083), Escherichia coli (ATCC 25922), and Candida albicans (ATCC-10231). The studies of doping and molecular dynamics were performed with the molecule that presented the highest concentration of drug–target proteins, 1IYL (C. albicans), 1C14 (E. coli), 2WE5 (E. faecalis), and 4TQX (S. mutans). The main compounds identified were: Curzerene (7.1%), γ-Elemene (7.04%), Germacrene D (7.61%), trans-β-Elemenone (11.78%), and Atractylone (18.65%). Gram positive bacteria and fungi were the most susceptible to the effects of the essential oil. The results obtained in the simulation showed that the major compound atractylone interacts with the catalytic sites of the target proteins, forming energetically favourable systems and remaining stable during the period of molecular dynamics.


Introduction
Fungi and bacteria can cause various pathologies in humans. Leprosy [1], tuberculosis [2], bacterial dysentery [3], gonorrhea [4], urinary tract infection, endocarditis [5,6], onychomycosis [7], mucormycosis [8], and candidiasis [9] are examples of diseases that these microorganisms can cause. In some cases, the symbiosis between bacteria and fungi increases the virulence of bacteria, because fungi such as C. albicans elevates the production of exopolysaccharides, which can become an ideal shelter for S. mutans, thus making it difficult to control this microorganism [10]. Another important factor is the resistance that microorganisms are developing to traditional antibiotics, since this poses a threat to public health and is associated with high rates of morbidity and mortality [11]. In this sense, natural products, more specifically essential oils, can become a viable alternative for the control of fungi and bacteria [12,13].
Siparuna guianensis was the first Siparuna species described and illustrated by Aublet [33]. This plant is present from Nicaragua to Paraguay, and in Brazil, this species is known by several names, such as negramina, folha-santa, marinheiro, capitiú, mata-cachorro, catingoso, limão-bravo, cicatrizante-das-guianas, catingueira-de-paca, and fedegoso. In many countries of America, leaves of S. guianensis are widely used as a drink to combat stomach pains [34] and this activity may be related to the compounds present in its essential oil [35,36]. In addition, there is little research that reports on the antimicrobial activity of Siparuna guianensis essential oil, [37] including a chemotype found in Tocantins, Brazil. In this context, the objective of this work was to evaluate the chemical composition, antimicrobial activity and simulate the mechanisms of interaction of the major chemical constituent present in the essential oil of Siparuna guianensis, using doping techniques and molecular dynamics.
In a study related to the chemical composition of S. guianensis essential oil from southeastern Brazil, they obtained high concentrations of capric acid (46.6%) and 2-undecane (31.7%) [41]. These compounds were not identified in other studies such as Zoghbi et al. [42], who analysed the chemical composition of S. guianensis essential oil, collected in various cities of Northern Brazil and identified epi-α-bisabolol (25.1%) and spathulenol (15. and atractylon (31.4%) in Belém (PA). The results found in other studies [43][44][45] show that the chemical composition of the essential oil of S. guianensis varies according to the seasonality and site of collection. results found in other studies [43][44][45] show that the chemical composition of the essential oil of S. guianensis varies according to the seasonality and site of collection.
In general, Gram-positive bacteria were the most sensitive to the effects of essential oil (EO), and this may be related to the fact that Gram-positive bacteria are more susceptible to the effects of volatile components compared to the Gram-negative ones [48]. In the case of fungi, EOs can be a viable alternative in the fight against the infection caused by Candida [24,49]. These biological effects can be related to the presence of chemically active compounds such as γ-elemene, curzerene, germacrene D, β-elemenone, and atractylon, as there are reports in the literature that corroborate this thesis [50][51][52]. [46] 0.51 32
In general, Gram-positive bacteria were the most sensitive to the effects of essential oil (EO), and this may be related to the fact that Gram-positive bacteria are more susceptible to the effects of volatile components compared to the Gram-negative ones [48]. In the case of fungi, EOs can be a viable alternative in the fight against the infection caused by Candida [24,49]. These biological effects can be related to the presence of chemically active compounds such as γ-elemene, curzerene, germacrene D, β-elemenone, and atractylon, as there are reports in the literature that corroborate this thesis [50][51][52]. [

Antimicrobial Activity
The antimicrobial activity analysed by the diffusion method can be observed in Table 2. The microorganisms presented mean inhibition halos of 11 ± 0.12 (mm), 12 ± 0.57(mm), 11 ± 0.31(mm), and 12.5 ± 0.98 (mm) for Gram-positive Streptococcus mutans (ATCC 3440), Gram-positive Enterococcus faecalis (ATCC-4083), Gram-negative Escherichia coli (ATCC 25922), and Candida albicans (ATCC-10231), respectively. Streptococcus mutans (ATCC-3440) and Candida albicans (ATCC-10231) were the most sensitive to the effects of essential oils, with a minimum inhibitory concentration of 125 µL/mL, whereas the bacterium Enterococcus faecalis (ATCC-4083) [36] demonstrated that the essential oil of S. guianensis exerts an inhibitory effect on fungi, and on Gram-negative and Gram-positive bacteria. In general, Gram-positive bacteria were the most sensitive to the effects of essential oil (EO), and this may be related to the fact that Gram-positive bacteria are more susceptible to the effects of volatile components compared to the Gram-negative ones [48]. In the case of fungi, EOs can be a viable alternative in the fight against the infection caused by Candida [24,49]. These biological effects can be related to the presence of chemically active compounds such as γ-elemene, curzerene, germacrene D, β-elemenone, and atractylon, as there are reports in the literature that corroborate this thesis [50][51][52].

Molecular Binding Mode
From our molecular docking results, it can be suggested that the ligand interacts favourably with the target proteins. In Table 3, the results of the MolDock score for each complex formed are presented. The interactions between atractylon, the primary compound from the essential oil isolated from the leaves of Siparuna guianensis, and the catalytic site of the enzymes were analysed. The interactions that were formed are visualized in Figure 2.  In Figure 2A, it is possible to observe that the ligand performed several hydrophobic interactions with different residues of the catalytic site of N-myristoyltransferase (C. albicans). With the Tyr225 residue, two interactions were established, one of the pi-pi types and the other of the pi-alkyl type. Residues Phe339 and Tyr354 had pi-alkyl-type interactions with the ligand, whereas Leu394 established alkyl interactions. Phe117 was also able to form two interactions, both of the pi-alkyl type. In Figure 2A, it is possible to observe that the ligand performed several hydrophobic interactions with different residues of the catalytic site of N-myristoyltransferase (C. albicans). With the Tyr225 residue, two interactions were established, one of the pi-pi types and the other of the pi-alkyl type. Residues Phe339 and Tyr354 had pi-alkyl-type interactions with the ligand, whereas Leu394 established alkyl interactions. Phe117 was also able to form two interactions, both of the pi-alkyl type.
With Enoyl reductase residues (E. coli), atractylon established six hydrophobic interactions. Four of these interactions were of the alkyl type with the following residues: Met206, Met159, Lys163, and Ala196. In addition, two additional pi-alkyl-type interactions with Tyr156 were formed.
The interaction of the ligand with the binding pocket of the enzyme Carbamate kinase (E. faecalis) can be seen in Figure 2C. With residues Val231, Cys235, and Met268, hydrophobic interactions of the alkyl type were formed. With Tyr238, two interactions were established, one of the pi-alkyl type and one of the pi-pi type. With Ala264, an interaction of the same type was formed.
All interactions formed with the residues of Sortase A (S. mutans) were of hydrophobic and alkyl types. These interactions were established with the following residues: Ile215, Val190, Ile191, Val188, Arg213, and Val2013.

Analysis of Complexes Stability
The complexes obtained by molecular docking were used as a starting point for molecular dynamics simulations. The root mean square deviation (RMSD) graphs were plotted in relation to the lowest energy structure obtained for the systems, after the execution of the protocol of energy minimization, heating, and equilibrium. To plot the RMSD of the proteins' backbone, their Cα atoms were used and to plot the RMSD of the ligands, their heavy atoms were used. The correspondent graphs can be seen in Figure 3.  For the systems formed with the target proteins of C. albicans, E. coli, E. faecalis and S. mutans, the mean RMSD obtained for the ligand was 0.65 Å, 0.62 Å, 0.66 Å, and 0.64 Å, respectively. Thus, it is possible to infer that during the simulations, the inhibitor remained stable at the binding site of the different targets during molecular dynamics.
The target proteins showed small conformational changes as can be observed in the RMSD plots. For the systems formed with the target proteins of C. albicans, E. coli, E. faecalis and S. mutans, the mean RMSD obtained for the ligand was 0.65 Å, 0.62 Å, 0.66 Å, and 0.64 Å, respectively. Thus, it is possible to infer that during the simulations, the inhibitor remained stable at the binding site of the different targets during molecular dynamics.
The target proteins showed small conformational changes as can be observed in the RMSD plots. These changes resulted from the accommodation of the ligands at their respective binding sites.
The fluctuations observed in the RMSD for the proteins backbone may be the result of the accommodation of the ligand at the active site. The mean values for the RMSD were relatively low. These values for the backbone of the target proteins of C. albicans, E. coli, E. faecalis, and S. mutans were 1.63 Å, 1.53 Å, 1.44 Å, and 1.65 Å, respectively.  (Table 4). In all systems, the free energy values demonstrated that atractylon is capable of inhibiting enzymatic activity. The contributions of Van der Waals were the most responsible for the interaction of the ligand with the molecular targets. Moreover, the electrostatic and nonpolar contributions were favourable for the maintenance of the complexes.

Preparation and Characterization of the Siparuna guianensis Sample
The Siparuna guianensis sample was obtained in the herbarium of the Museu Paraense Emilio Goeld (Eastern Amazon), on 09/09/2016. The geographical coordinates of the collection site were S01 • 27 04.3" and W048 • 26 38.3", with a relative humidity of 64.9% and temperature of 26.5 • C. The samples were identified by Dr. Antonio Elielson Sousa da Rocha, and then, its registration number was incorporated in the Emílio Goeldi Museum Herbarium, located in the city of Belém, Pará, Brazil, under the v-oucher MG-165435. Before the extraction process, the sample was dried and ground and then the moisture content was determined by infrared moisture analyser. The images of the leaves of S. guianensis can be observed in Figure 4. S01°27'04.3'' and W048°26'38.3'', with a relative humidity of 64.9% and temperature of 26.5 °C. The samples were identified by Dr. Antonio Elielson Sousa da Rocha, and then, its registration number was incorporated in the Emílio Goeldi Museum Herbarium, located in the city of Belém, Pará, Brazil, under the v-oucher MG-165435. Before the extraction process, the sample was dried and ground and then the moisture content was determined by infrared moisture analyser. The images of the leaves of S. guianensis can be observed in Figure 4. It is a shrub about three meters high, the immature fruits were greenish, and the ripe ones were greenish and purplish, with short pedunculated axillary racemes, and opposite, elliptic, and lanceolate leaves. This plant releases a characteristic odour of fish. Leaf samples were identified and deposited in the medicinal plants herbarium of the Museu Paraense Emílio Goeldi.

Extraction Procedure: Hydrodistillation
After the drying process, the leaves of S. guianensis were submitted to hydrodistillation using a Clevenger-type extractor. For the extraction process, 40g of the sample was used, for 10,800 s at 100 °C. After this procedure, anhydrous sodium sulphate (Na2SO4) was added and the essential oil was centrifuged to be moisture-free. The essential oil yield was calculated in dry basis (db).

Analysis of Volatile Compounds
The chemical composition of the essential oils was evaluated by gas chromatography/mass spectrometry (GC/MS) according to the methodologies by [53,54], using a Shimadzu QP-2010 plus system under the following conditions: silica capillary column Rtx-5MS (30 m × 0.25 mm, 0.25 μm film thickness); program temperature of 60-240 °C at 3 °C/min; injector temperature of 250 °C; helium as carrier gas (linear velocity of 32 cm/s, measured at 100 °C); splitless injection (1 μL of a 2:1000 hexane solution). Ionization was obtained by electronic impact technique at 70 eV, and the temperature of the ion source and other parts was set at 200 °C. The quantification of volatile compounds was determined by gas chromatography with a flame ionization detector (FID; Shimadzu, QP 2010 system-Kyoto, Japan) under the same conditions as gas chromatography coupled It is a shrub about three meters high, the immature fruits were greenish, and the ripe ones were greenish and purplish, with short pedunculated axillary racemes, and opposite, elliptic, and lanceolate leaves. This plant releases a characteristic odour of fish. Leaf samples were identified and deposited in the medicinal plants herbarium of the Museu Paraense Emílio Goeldi.

Extraction Procedure: Hydrodistillation
After the drying process, the leaves of S. guianensis were submitted to hydrodistillation using a Clevenger-type extractor. For the extraction process, 40 g of the sample was used, for 10,800 s at 100 • C. After this procedure, anhydrous sodium sulphate (Na 2 SO 4 ) was added and the essential oil was centrifuged to be moisture-free. The essential oil yield was calculated in dry basis (db).

Analysis of Volatile Compounds
The chemical composition of the essential oils was evaluated by gas chromatography/mass spectrometry (GC/MS) according to the methodologies by [53,54], using a Shimadzu QP-2010 plus system under the following conditions: silica capillary column Rtx-5MS (30 m × 0.25 mm, 0.25 µm film thickness); program temperature of 60-240 • C at 3 • C/min; injector temperature of 250 • C; helium as carrier gas (linear velocity of 32 cm/s, measured at 100 • C); splitless injection (1 µL of a 2:1000 hexane solution). Ionization was obtained by electronic impact technique at 70 eV, and the temperature of the ion source and other parts was set at 200 • C. The quantification of volatile compounds was determined by gas chromatography with a flame ionization detector (FID; Shimadzu, QP 2010 system-Kyoto, Japan) under the same conditions as gas chromatography coupled to mass spectrometry (GC-MS), except that hydrogen was used as the carrier gas. The retention index was calculated for all volatile constituents using a homologous series of n-alkanes (C 8 -C 20 ), and were identified by comparing the mass spectra obtained experimentally and their retention indices to those found in literature [46,47].

Analysis of In vitro Antimicrobial Activity
In the microbiological assays, standard strains of Streptococcus mutans (ATCC 3440), Enterococcus faecalis (ATCC 4083), Escherichia coli (ATCC 25922), and Candida albicans (ATCC 10231) were used. All of them were purchased from the Osvaldo Cruz Foundation (FIOCRUZ-Rio de Janeiro, Brazil), belonging to the base of standards of the Laboratory of Microbiological Quality Control of Medicines of the University Center of Pará-CESUPA.
The inoculum of each microorganism was obtained from a microbial suspension of fresh culture (maximum 24 h) in saline solution 0.85% (m/V), by comparing the inoculum turbidity with the MacFarland scale, equivalent to a concentration of 1.5 × 108 UFC/mL [55] in a turbidimeter (Grant bio, Model: DEN-1Shepreth, Cambridge, UK).
The culture medium used for the disk diffusion test was soybean casein agar (SCA) and brain Heart Infusion (BHI) broth containing 0.2% polysorbate 80 (m/V). 5% (v/v) of sheep blood was added for the analysis of strains of Streptococcus mutans (ATCC 3440) and Enterococcus faecalis (ATCC 4083).

Evaluation of the Sample Sensitivity by the Disk Diffusion Method
Briefly, 10 mL of Soybean Casein Agar (15 × 100 mm) was poured into a Petri dish. The microorganism (106 CFU/mL) was then inoculated with the aid of a sterile swab and paper discs (6 and 8 m) impregnated with 10 µL of oil. Positive and negative control were added onto the medium. The plates were incubated at 30 ± 5 • C/24 h in an aerobic environment [22,55,56]. The analysis was performed in triplicate. After the incubation period, the plates were revealed with triphenyltetrazolic chloride at 7 mg/mL in bacteriological agar at 1% (w/v). The halos were measured using a pachymeter (mm) and evaluated by a descriptive analysis.

Determination of Minimum Inhibitory Concentration (MIC)
The Minimum inhibitory concentration MIC was performed with the essential oil and was adapted from the micro dilution proposed by [55]. The test was performed on an Elisa ® plate, where a 100 µL sample aliquot was diluted (1:2 v/v) in BHI broth containing 106 CFU/mL until 10 consecutive dilutions, and then positive and negatives controls were added. Plates were incubated at 30 ± 5 • C/48 h. The test was performed in triplicate.
After incubation, plates were revealed with 1% (m/v) bacteriological broth containing 7 mg/mL triphenyltetrazolic chloride solution and incubated for further 30 min at 30 ± 5 • C for bacteria, and at 25 ± 5 • C for C. albicans. The maintenance of the red colour in the medium was interpreted as microbial growth.

Statistical Analysis
The statistical analysis was performed using the application MiniTab17-State College, Pennsylvania, USA, using the means and their respective standard deviations.

Molecular Docking
For the study of molecular docking, atractylone was selected, since it was the primary compound from the essential oil isolated from the leaves of Siparuna guianensis. Molecular docking was used to investigate the interaction between atractylone and essential proteins of C. albicans, E. coli, E. faecalis, and S. mutans. The proteins used as a molecular target are essential for the metabolic pathways of such microorganisms, in addition to being reported in the literature as targets for natural and synthetic products that combat these pathogens [57][58][59][60].
The chemical structure of atractylon, after being designed with GaussView 5.5 software -Wallingford, Connecticut United States, was optimized with B3LYP/6-31G* [61], using Gaussian 16 (Wallingford, CT, USA) [62]. To study the interaction mode of this molecule with target-proteins for drug action, the software Molegro Virtual Docker 6 (Århus, Denmark, https://molegrovirtualdocker.weebly.com/) was used [63,64]. The crystal structures of the proteins used as targets can be found in the Protein Data Bank (www.rcsb.org), from their ID: 1IYL (C. albicans) [57], 1C14 (E. coli) [58], 2WE5 (E. faecalis) [59], and 4TQX (S. mutans) [60]. The MolDock Score (GRID) function was used with a Grid resolution of 0.30 Å and radius of 7 Å, encompassing the entire crystallographic ligand binding cavity found in the PDB of each protein. The MolDock SE algorithm was used with number of runs equal to 10; 1500 max interactions, and max population size equal to 50. The maximum evaluation of 300 steps with neighbor distance factor equal to 1 and energy threshold equal to 100 were used during the molecular docking simulation. The RMSD limit for multiple cluster poses was set to < 1.00 Å.

Molecular Dynamics (MD) Simulation
The ligand parameters were constructed with the aid of the Antechamber module, using the General Amber Force Field (GAFF) [65]. The calculations to determine the atomic charges of the ligand were performed according to the restrained electrostatic potential (RESP) protocol using basis set Hartree-Fock level might with the functional 6-31G* [66]. To measure the protonation status of the amino acid residues of the receptors, the results obtained from the PROPKA program were used [67,68].
In the molecular dynamics simulations, the force field ff14SB [69] and the explicit water molecules described by the TIP3P model [70] were used. All systems were solvated in an octahedron periodic box, where a cutting radius of 12Å was applied in all directions from the solute. Finally, in each system, an adequate number of counter-ions were added to neutralize the charge.
The MD simulations were performed with the Amber 16 package [71][72][73]. Sander. MPI was used for the energy minimization steps, and pmemd. CUDA, for the heating, equilibrium and MD simulations.
The energy minimization of the systems occurred in three stages. In the first step, 1500 cycles were performed using steepest descent method and conjugate gradient algorithm, applying a harmonic force constant of 100 kcal/mol.Å −2 on the solute. In the second step, the harmonic force constant applied on the solute was 50 kcal/mol.Å −2 and further 500 cycles were performed using the steepest descent method and conjugate gradient algorithm. In the last step, the restrictions were removed, and 500 cycles were performed using the same protocol.
To raise the systems temperature from 0 to 300k, 800 ps of simulations were performed. The heating was carried out in three stages. In the first stage, the solute was restricted with a harmonic force constant of 50 kcal/mol.Å −2 . Thus, only the solvent and the counter-ions get free to move. In the next two steps, the harmonic force constant was removed.
To balance the complexes, 2 ns of simulations with constant temperature and with no restrictions were performed. Then, for each complex, 100 ns of MD simulation were obtained with NVT ensemble.
The particle mesh Ewald method [74] was used for the calculation of electrostatic interactions, and the bonds involving hydrogen atoms were restricted with the SHAKE algorithm [75]. The temperature control was performed with the Langevin thermostat [76] within collision frequency of 2 ps −1 .

Free Energy Calculations
The binding free energy was calculated using the molecular mechanics-generalized Born surface area (MM-GBSA) approach [77][78][79]. For the affinity energy calculation, 500 snapshots of the last 5 ns of the MD simulations trajectories were used.
The free energy was calculated according to the following equations: where ∆G bind is the free energy of the complex, which is the result of the sum of the molecular mechanics energy (∆E MM ), the desolvation free energy (∆G solv ), and the entropy (−T∆S).
The energy of molecular gas phase mechanics (∆E MM ) can be described by the sum of the internal energy contributions (∆E internal ), the sum of the connection, angle and dihedral energies, electrostatic contributions (∆E eletrostatic ), and van der Waals terms (∆E vdW ).
Desolvation free energy (∆G solv ) is the sum of the polar (∆G GB ) and non-polar (∆G nonpol ) contributions. The polar desolvation term was calculated using the implicit generalized born (GB) approaches.