An In Vitro and In Silico Investigation about Monteverdia ilicifolia Activity against Helicobacter pylori

Monteverdia ilicifolia is a Brazilian native plant, traditionally used to treat gastric diseases that are now associated with Helicobacter pylori and are commonly associated with several human diseases. We point out the M. ilicifolia extract as active against H. pylori. The crude extract produced with acetone:water presented the best H. pylori inhibitory activity of all five extracts (MIC 64 µg/mL). The ethyl-acetate fractions from crude extracts produced with ethanol and acetone showed a MIC of 64 µg/mL. Both ethyl-acetate fractions and the crude extract produced with acetone showed an antioxidant capacity of between 14.51 and 19.48 µg/mL in the DPPH assay. In the FRAP assay, two ethyl-acetate fractions (EAF2 and EAF4) presented the antioxidant capacity of 5.40 and 5.15 mM Trolox/g of extract. According to the results obtained from the antioxidant and antibacterial assays, two fractions (EAF2 and nBF5) were analyzed by mass spectrometry and confirmed the presence of monomeric, dimeric, trimeric tannins, and glycosylated flavonoids. Some compounds were tested using bioinformatics to evaluate the best enzyme inhibitors and the molecular interaction between the enzyme and the tested ligands. The presence of these polyphenol compounds could play an important role in antioxidant and inhibitory capacities against H. pylori and can be used to assist in the treatment or prevention of infection by H. pylori.


Introduction
Helicobacter pylori is a Gram-negative bacillus that can contribute to the development of human diseases, such as gastric ulcers, gastritis, gastric adenocarcinoma, and lymphoma. The bacteria express some virulence factors that expand the possibilities of the bacteria interacting with the host cells. One of those factors allows the bacteria to neutralize the acidity of gastric secretion, hydrolyzing urea to ammonium and CO 2 , promoting the adjustment of pH in the stomach to neutral, and producing a cytopathic effect in the stomach cells. This mechanism allows the bacteria to survive in stomach conditions [1,2].
In 2015 approximately 50% of the world's population (4.4 billion) was estimated to be infected with H. pylori. In developing and newly industrialized countries, the prevalence is higher than in developed countries. This difference reflects the level of urbanization, sanitation, access to clean water, and varied socioeconomic statuses. Brazil has one of the highest infection prevalences in Latin America and the Caribbean (71.2%) [3].
The first-line treatment for H. pylori infection is one proton pump inhibitor, amoxicillin, and clarithromycin for 14 days. As an alternative, treatment can be used in quadruple

Epicatechin Determination Using HPLC
According to the Brazilian Pharmacopeia [19], the M. ilicifolia plant drug must present at least 2.8 mg/g in the extract of epicatechin. All extracts prepared showed an epicatechin concentration higher than the preconized limit. The profile for the five extracts is presented in Figure 1. The CE1 presented the lower epicatechin concentration (mg/g of extract) of 3.6 ± 0.02, followed by the CE2 at 15.54 ± 0.14, CE4 at 16.77 ± 2.14, and CE3 at 19.63 ± 2.04. The CE5 presented a higher epicatechin concentration of 20.66 ± 1.99.

Antibacterial Activities of Extracts and Semi-Purified Fractions against H. pylori
Initially, the effect of the crude extracts, ethyl-acetate, n-butanolic, and aqueous fractions was investigated at different concentrations (32-1024 µg/mL), as shown in Table 1.

Antibacterial Activities of Extracts and Semi-Purified Fractions against H. pylori
Initially, the effect of the crude extracts, ethyl-acetate, n-butanolic, and aqueous fractions was investigated at different concentrations (32-1024 µ g/mL), as shown in Table 1. The best inhibitory activity corresponded to the CE5, EAF2, and EAF5 with an MIC50 of 64 µ g/mL and nBF5 of 128 µ g/mL. It was observed that partition with ethyl acetate demonstrated better MIC in comparison with other fractions.
The antibacterial result obtained is concordant with the result found in a previous study using Plumbago zeylanica L. The extract produced with acetone presented the lowest MICs when compared with the ethanol:water extract. In another study performed with Sclerocarya birrea (A.Rich.) Hochst, the result was similar, where the extract produced with acetone:water showed a MIC comparable with amoxicillin and metronidazole [20,21]. The best inhibitory activity corresponded to the CE5, EAF2, and EAF5 with an MIC 50 of 64 µg/mL and nBF5 of 128 µg/mL. It was observed that partition with ethyl acetate demonstrated better MIC in comparison with other fractions.
The antibacterial result obtained is concordant with the result found in a previous study using Plumbago zeylanica L. The extract produced with acetone presented the lowest MICs when compared with the ethanol:water extract. In another study performed with Sclerocarya birrea (A.Rich.) Hochst, the result was similar, where the extract produced with acetone:water showed a MIC comparable with amoxicillin and metronidazole [20,21].
The fractions produced with the ethyl acetate and n-butanol for the ethanol:water 50:50 (v/v) and acetone:water 7:3 (v/v) extracts showed to be MIC between 64 and 256 µg/mL. This can be explained because when a crude extract is prepared using the mixture of acetone:water or when a fraction is obtained by the partition of crude extracts with ethyl acetate, the compounds that are mostly obtained in both cases are phenolic compounds [22].
The result obtained for the nBF inhibitory activity (128 µg/mL) can be compared with the result shown in the study using Rosa hybrida Colorado extract that is, it is rich in polyphenols compounds when the MIC was 10 µg/mL to the butanolic fraction and 100 µg/mL to the ethanolic fraction [23].
One of H. pylori's virulence factors is the presence of the urease enzyme. This enzyme is responsible for allowing the bacteria to stay alive in the acid stomach environment [1,2]. The ability of M. ilicifolia extracts and fractions to inhibit the bacterial urease enzyme was evaluated. The results (Table S1) showed that nBF5 presented the best inhibition (47.08%), followed by EAF2 (40.50%). nBF2 and EAF5 showed an enzyme inhibition of 37.27% and 37.51%, respectively.
The anti-H. pylori activity was tested for several traditional herbs in past studies and showed that Cimicifuga heracleifolia Kom could inhibit the urease enzyme by 42% [24]. It was related to the ability of several phenolic compounds to inhibit the urease enzyme. Compounds that had catechol as a skeleton showed the most potent inhibitory activities, ranging from 38 to 94%. This activity is attributed to the two ortho-hydroxyl groups that are presented in the aromatic ring of polyphenols molecules [25]. According to Pessuto [11], the number of polyphenolic hydroxyls present in the compound structure and the stereochemistry of the compounds are directly related to the ability to scavenge free radicals.
M. ilicifolia is a plant rich in polyphenols compounds, such as (epi)catechin (I),  (Figure 2). Due to the well-known presence of phenolic compounds in the specie, we can suggest that the anti-H. pylori activity observed is a result of their presence [11,26,27].

UPLC-MS Profiles of M. ilicifolia Extracts
Based on the anti-H. pylori assays, two fractions were chosen to follow the chemistry characterization. EAF2 and nBF5 were used in the UPLC-MS analysis. Table 2 shows the retention time of the compounds separated for chromatography from EAF2, as well as the ion and correspondent fragment in the negative mode.

UPLC-MS Profiles of M. ilicifolia Extracts
Based on the anti-H. pylori assays, two fractions were chosen to follow the chemistry characterization. EAF2 and nBF5 were used in the UPLC-MS analysis. Table 2 shows the retention time of the compounds separated for chromatography from EAF2, as well as the ion and correspondent fragment in the negative mode.
In the nBF5 fraction, it was characterized by glycosylated flavonoids such as kaempferolgalactoside-rhamnoside-rhamnoside and quercetin-rhamnopiranosyl-glucopiranosiderhamonoside; besides this, the fraction presents some compounds that could not be identified (Table S2).
Several studies have shown the presence of tannins and flavonoids in M. ilicifolia extracts that are considered responsible for their biological activity. Monomeric and dimeric flavonoids, such as epicatechin and procyanidins B1 and B2, were isolated and charac-terized in studies over the years from aqueous, hexanic, and acetonic extracts [11,28,29]. Glycosylated flavonoids are also present in the leaves extracted from M. aquifolium and M. ilicifolia and are composed of quercetin and kaempferol 3-O-glycosides [27,28,30,31]. Some authors have already demonstrated that procyanidins, catechin, and gallic acid isolated from natural products can have activity against H. pylori with reference strain and antibiotic-sensitive and resistant clinical isolates [32][33][34][35][36]. All these studies support the characterization proposed in the present work, considering that the same substances were isolated and characterized in those studies carried out with the specie. With the results obtained in the activity assays conducted with the fractions, we can presume that the phenolic compounds are responsible for the activity, as several authors demonstrated [11,27,28,30,31].

Antioxidant Capacity of Extracts and Semi-Purified Fractions
For the antioxidant capacity assays, seven extracts and fractions were selected that showed the best activity against H. pylori. The DPPH results are shown in Figure S1; the antioxidant capacity is present as IC 50 , which means that the extract concentration is needed to promote 50% of the antioxidant activity.
The antioxidant capacity for the tested extracts varies between 14.51 and 98.35 µg/mL. The more pronounced antioxidant capacity was observed in EAF4 with an IC 50 14.51 µg/mL followed by CE5 (IC 50 19.08 µg/mL) and EAF2 (IC 50 19.48 µg/mL), with no significant statistical difference. The worst antioxidant capacity in the DPPH assay was observed for nBF5 with an IC 50 98.35 µg/mL and the AQF3 (IC 50 94.72 µg/mL). The quercetin, used as a positive control, presented an IC 50 of 2.99 µg/mL. It is known that the antioxidant capacity is more pronounced in the presence of polyphenols [11]. This can explain why the extract obtained with acetone and water was rich in phenolic compounds and had a more notable result.
A previous study tested the antioxidant capacity using the DPPH radical scavenging method of the ethyl-acetate and n-butanolic fraction from M. royleana (Wall. ex M.A. Lawson) Cufod, both rich in polyphenols. The ethyl-acetate fraction showed an IC 50 of 55.01 µg/mL, while the n-butanolic fraction was 58.01 µg/mL. The antioxidant capacity was evaluated for the ethyl-acetate fraction from an acetonic crude extract by the DPPH radical scavenging method and showed a result of 25.39 µg/mL [11,37].
For the FRAP assay, the values of the antioxidant capacity were expressed as an mM Trolox/g extract equivalent ( Figure S2). In this assay, the absorbance variance found was linearly proportional to the antioxidant concentration.
The FRAP assay showed results varying from 0.77 to 5.40 mM Trolox/g of the extract. The EAF2 and EAF4 showed the result of 5.40 and 5.15 mM Trolox/g of the extract, respectively, with no significant statistical difference. The quercetin was used as the positive control, and the antioxidant capacity was 15.41 mM Trolox/g of the extract.
Some species from the Celastraceae family were tested to determine the antioxidant capacity using the FRAP assay. They found Cassine orientalis (Jacq.) Kuntze e M. pyria (Willemet) N. Robson has an antioxidant capacity of 584 ± 5.24 e 190 ± 0.87 µM trolox/g in the extract [38].
Polyphenols have a great chemical structure for the removal of free radicals. For the best activity, these compounds must present some specific structural characteristics with the hydroxyl groups and in the ring substitution. Some structural specifications were tested previously, and it was confirmed that the presence of the ortho-hydroxyl group in the ring B increases the antioxidant capacity [39,40].
The compounds present in the extracts produced with organic solvents can eliminate free radicals more effectively than those compounds with more polarity present in aqueous extracts or fractions; the explanation could be the higher presence of phenolic compounds in the extracts produced using organic solvents.
The fact that FAE2 showed the best antioxidant capacity over the nBF could be explained by the fact that EAF2 is richer in phenolic compounds. These results are according to the conclusions of [11], who said that phenolic hydroxyl present in the phenolic compounds, as the fraction with the best antioxidant capacity and the one rich in phenolic compounds with orto-hydroxyl in the structure, are more capable of scavenging free radicals.
In summary, the present study demonstrated the anti-H. pylori activity of M. ilicifolia extracts and fractions represents a strong potential for use in the treatment or even more strongly acting in the prevention of H. pylori infection. The antioxidant capacity was confirmed for the specie using two different methods. Both activities can be attributed to the presence of phenolic compounds, such as monomeric, dimeric, and glycosylated flavonoids, which can be found in higher amounts in the extracts and fractions produced with organic solvents.

Virtual Screening
To identify the compound most likely to act as the H. pylori urease inhibitor from those present in the M. ilicifolia extract, four molecular docking simulations for each compound in the library using two different programs were carried out. In this way, the mean scores obtained from each program were used in the calculation of the mean relative score expressed by Equation (1). The compound kaempferol-3-galactoside-6-rhamnoside-3rhamnoside, named CID 44258967, had the highest mean relative score, followed by the compound (epi)afzelechin-(epi)catechin-(epi)catechin, named AFZ-CAT-CAT ( Figure 3). For these compounds, the best pose obtained in each docking simulation in the Gold program showed a recurrent conformation pattern in the urease active site ( Figure S3), which suggests a site-specific interaction profile with the enzyme, which is characteristic of compounds with a drug-like behavior. Together, these data suggest that the compounds CID 44258967 and AFZ-CAT-CAT are the most likely urease inhibitors presented in the M. ilicifolia extract.
To describe the molecular interactions that occur between urease and the best MRS ligands, the PoseView program [41] was used. For the compound CID44258967 ( Figure 4A), the program predicts the hydrogen bonds with residues His221, Glu222, Thr251, Ala278, and Arg338, hydrophobic contacts with the residues Met317, Leu318, Cys321, and Phe334, a π-π stacking with residue Phe334, and a charge-dipole interaction with one of the Ni 2+ cofactors. For the compound AFZ-CAT-CAT ( Figure 4B), hydrogen bonds are predicted with the residues Ala169, Glu222, Asp223, and Asp362, a hydrophobic contact with Met366 and charge-dipole interaction with the two Ni 2+ cofactors.
There are no reports in the literature citing the activity of kaempferol-3-galactoside-6-rhamnoside-3-rhamnoside or (epi)afzelechin-(epi)catechin-(epi)catechin as urease inhibitors, which makes this work the first one to associate anti-ureolytic activity with these two compounds. To describe the molecular interactions that occur between urease and the best MRS ligands, the PoseView program [41] was used. For the compound CID44258967 ( Figure 4A), the program predicts the hydrogen bonds with residues His221, Glu222, Thr251, Ala278, and Arg338, hydrophobic contacts with the residues Met317, Leu318, Cys321, and Phe334, a π-π stacking with residue Phe334, and a charge-dipole interaction with one of the Ni 2+ cofactors. For the compound AFZ-CAT-CAT ( Figure 4B), hydrogen bonds are predicted with the residues Ala169, Glu222, Asp223, and Asp362, a hydrophobic contact with Met366 and charge-dipole interaction with the two Ni 2+ cofactors. There are no reports in the literature citing the activity of kaempferol-3-galactoside-6-rhamnoside-3-rhamnoside or (epi)afzelechin-(epi)catechin-(epi)catechin as urease Regarding toxicity, as far as we know, there are no reports in the literature describing the possible toxic effects of these two compounds, which leads us to evaluate their toxicity in silico by the SwissADME server. As a result, the AFZ-CAT-CAT compound showed three violations of Lipinski's rules, MW > 500, the number of N or O > 10, and the number of NH or OH > 5. In addition, it has an alert as a possible pan assay interference compound (PAIN). The compound CID44258967 presented the same three violations of Lipinski's rules but no alert as PAIN. However, Lipinsky's rules assess the usability of the compounds as oral drugs, not regarding their toxic effects. In this way, the extract, and fractions of M. ilicifolia used in this work, have already been evaluated for toxicity and showed no relevant effect on the mitochondrial activity of human stomach AGS cells (AGS, ATCC CRL-1739) [42]. To describe the molecular interactions that occur between urease and the best MRS ligands, the PoseView program [41] was used. For the compound CID44258967 ( Figure 4A), the program predicts the hydrogen bonds with residues His221, Glu222, Thr251, Ala278, and Arg338, hydrophobic contacts with the residues Met317, Leu318, Cys321, and Phe334, a π-π stacking with residue Phe334, and a charge-dipole interaction with one of the Ni 2+ cofactors. For the compound AFZ-CAT-CAT ( Figure 4B), hydrogen bonds are predicted with the residues Ala169, Glu222, Asp223, and Asp362, a hydrophobic contact with Met366 and charge-dipole interaction with the two Ni 2+ cofactors. There are no reports in the literature citing the activity of kaempferol-3-galactoside-6-rhamnoside-3-rhamnoside or (epi)afzelechin-(epi)catechin-(epi)catechin as urease

Plant Material
The leaves of M. ilicifolia were collected in Marialva, Brazil (23 •
The sample preparation was carried out following a previous study using 1 g of each crude extract instead the pulverized leaves. The standard solution's preparation and obtention of the calibration curve were conducted according to the previous study (Epicatechin calibration curve: y = 151,061x + 2,726,000) [29].

Bacteria Strain
It used H. pylori strain ATCC ® 43504, amoxicillin sensitive, and metronidazole resistance. Bacteria were grown in the Columbia Agar supplemented with sheep blood (5%) in a 10.0% CO 2 atmosphere at 37.0 • C for 72 h [43].

Minimum Inhibitory Concentration (MIC) and Minimum Bactericidal Concentration (MBC)
The anti-H. pylori assay was conducted by the microdilution method [44]. As a positive control, amoxicillin and metronidazole (Sigma Chemical Co., St. Louis, MO, USA) were used. In each well, 100 µL of the sample solutions (32-1024 µg/mL) and 100 µL H. pylori suspension (≈10 6 -10 7 bacteria/mL) were added both in supplemented BHI. The absorbance was measured at 620 nm and then incubated (37.0 • C/72 h/10.0% CO 2 ). After incubation, the plate was homogenized, and a new measurement was performed to determine the MIC.
The MBC assay was performed for samples that presented MIC. The sample corresponding to the microplate well without apparent growth in BHI was harvested in a Columbia Agar plate (5% sheep blood) and incubated at 37.0 • C, 10.0% CO 2 , 72 h. The assay was determined by the lowest sample concentration able to inhibit colony formation.

Urease Inhibition Assay
Urease inhibition activity was determined based on the production of ammonia catalyzed by the enzyme urease, according to the method described by [44]. The reaction microplate contained a mixture of 25 µL of urease 4 UI (Sigma Jack Bean urease type III) and 25 µL of the sample at varying concentrations, and it was incubated at room temperature for two hours. Then, 25 µL of phenol red (0.02%) and 200 µL of urea (50 mM) in 100 mM phosphate buffer (pH 6.8) were added to the microplate. After 20 min, the mixture absorbance was read at 540 mm using a microplate reader (iMark ® , BioRad, Washington, DC, USA). Boric acid was used as the standard positive control for urease inhibition.

UHPLC-MS Conditions
UHPLC-MS analysis was performed on a Nexera X2 liquid chromatography system with an LC-30AD pump and Phenomenex ® Gemini C-18 column (250 mm × 4.6 mm) coupled with a Q-TOF Impact II (Bruker Daltonics, Bremen, Germany), with electrospray ionization source. The column was maintained at 40 • C with a linear elution gradient of water 0.1% formic acid (eluent A) and acetonitrile 0.

Virtual Screening
One of the beta subunits of the hetero 24-mer urease from H. pylori and bonded to the inhibitor 2-{[1-(3,5-dimethylphenyl)-1H-imidazol-2-yl]sulfanyl}-N-hydroxyacetamide (named DJM; CID 42065735) (PDB id: 6ZJA) was used in the virtual screening of compounds described in the M. ilicifolia extract (Table S4). The programs and respective protocols were defined based on the redocking of the ligand in the urease. The AutoDock-Vina (Vina) program [46] used the standard search and ranking algorithms, with the search space defined by the box centered on the ligand, with dimensions of 30, 20, and 30 Å at x, y, and z, respectively. The uff force field [47] was used to minimize the ligands. The program Gold [48] uses the search space with a 15 Å radius centered on the ligand. The search algorithm employed a 200% efficiency, with 30 runs, and the ASP scoring and the 'Allow early termination' option was disabled.
The compounds described in the M. ilicifolia extract plus the reference ligand CID 42065735 were in the library for the virtual screening. The 2D or 3D structures of the compounds were obtained from the PubChem database or drawn by the ChemDraw ® program. Different isomers were used when necessary, and the OpenBabel program [49] was used to add hydrogens and convert them to 3D.
The mean relative score (MRS) was calculated from the mean docking score of each compound after four simulations using each program (Equation (1)). In this equation, Vina represents the average score of the compound obtained from four simulations using the Vina program, and Vina max expresses the maximum average score observed for the given compounds of the library; the same concept was applied to the results from the Gold program (Gold and Gold max ). The MRS allows the ranking of compounds with the ligand CID42065735 as a reference.
The in-silico toxicity was estimated by the SwissADME server [50].

Statistical Analysis
Numerical data were presented as the means ± standard deviation (SD). The number of repetitions of individual assays was different and was present in the description of each of them. One-way ANOVA with Bonferroni post-test was performed considering values of p ≤ 0.05 as statistically significant.

Conclusions
The EAF2 and nBF5 presented the best in vitro anti-H. pylori activity among all the extracts and fractions tested. In the urease enzyme inhibition test, the highest percentage of inhibition was observed in the nBF5 and EAF2. The antioxidant capacity was better observed in the ethyl acetate fractions tested and for the CE5. The in silico assays demonstrated that kaempferol-3-galactoside-6-rhamnoside-3-rhamnoside and (epi)afzelechin-(epi)catechin-(epi)catechin presented in the fractions EAF2 and nBF5 and have an anti-ureolytic activity. The anti-H. pylori and antioxidant activity observed can be attributed to the phenolic compounds present in EAF2 and nBF5.

Conflicts of Interest:
The authors declare no conflict of interest.

CE1
Crude extract aqueous EAF1 Ethyl-acetate fraction from crude extract aqueous nBF1 n-Butanolic fraction from crude extract aqueous AQF1 Aqueous fraction from crude extract aqueous CE2 Crude