Flavones from Combretum quadrangulare Growing in Vietnam and Their Alpha-Glucosidase Inhibitory Activity

Combretum quadrangulare Kurz is widely used in folk medicine in Eastern Asia and is associated with various ethnopharmacological properties including hepatoprotective, antipyretic, analgesic, antidysenteric, and anthelmintic activities. Previous phytochemical investigations reported the presence of numerous triterpenes (mostly cycloartanes, ursanes, lupanes, and oleananes) along with dozens of flavonoids. However, the extracts of C. quadrangulare and isolated flavonoids have not been evaluated for their alpha-glucosidase inhibition. In the frame of our efforts dedicated to the chemical investigation of Vietnamese medicinal plants and their biological activities, a phytochemical study of the MeOH extract of the leaves of C. quadrangulare using bioactive guided isolation was undertaken. In this paper, the isolation and structure elucidation of twelve known compounds, 5-hydroxy-3,7,4′-trimethoxyflavone (1), ayanin (2), kumatakenin (3), rhamnocitrin (4), ombuin (5), myricetin-3,7,3′,5′-tetramethyl ether (6), gardenin D (7), luteolin (12), apigenin (13), mearnsetin (14), isoorientin (15), and vitexin (16) were reported. Bromination was applied to compounds 2 and 3 to provide four new synthetic analogues 8–11. All isolated and synthesized compounds were evaluated for alpha-glucosidase inhibition and antibacterial activity. Compounds 4 and 5 showed moderate antibacterial activity against methicillin-resistant Staphylococcus aureus while others were inactive. All compounds failed to reveal any activity toward extended spectrum beta-lactamase-producing Escherichia coli. Compounds 2, 4, 6–9, and 11–14 showed good alpha-glucosidase inhibition with IC50 values in the range of 30.5–282.0 µM. The kinetic of enzyme inhibition showed that 8 and 11 were noncompetitive type inhibition against alpha-glucosidase. In silico molecular docking model indicated that compounds 8 and 11 were potential inhibitors against enzyme α-glucosidase.


Phytochemical Study and Derivatization of Compounds 2 and 3
The crude MeOH extract was successively partitioned into n-hexane, n-hexane:EtOAc (1:1, v/v), and EtOAc to provide extracts H, HEA, EA, and M, respectively. These fractions were evaluated for α-glucosidase inhibition. The most active extract was HEA. This was fractionated by column chromatography (CC), providing the fractions P1-P9. Of these, P2 and P9 showed the strongest biological activity (see Table S6). Isolation and purification were performed on these fractions (see Experimental Section), affording compounds 1-7 and 12-16.
Compounds 2 and 3 were selected for bromination to obtain new flavones (Scheme 1). The reactions were conducted using hydrogen peroxide and potassium bromide in acetic acid (Scheme 1). As a result, compound 8 was prepared from 2 while products 9-11 were derived from 3. The isolated yields of these compounds were 35-93%. The 1H NMR spectrum of 8 displayed two meta-coupled protons at δ H 7.88 and 8.18, three methoxy groups at δ H 3.80 and 3.97 (x2), and one hydrogen-bond hydroxyl group at δ H 15.71. NMR data of 8 and its mother compound 2 was highly similar. The differences are the disappearance of three aromatic protons (H-6, H-8, and H-5 ), indicating that three positions were brominated. Altogether, the chemical structure of 8 was elucidated as shown. The identification of compounds 9-11 was readily established on the basis of their 1 H, 13 C NMR, and HRESI mass spectra (Tables S3 and S4). Compounds 2 and 3 were selected for bromination to obtain new flavones (Scheme 1). The reactions were conducted using hydrogen peroxide and potassium bromide in acetic acid (Scheme 1). As a result, compound 8 was prepared from 2 while products 9-11 were derived from 3. The isolated yields of these compounds were 35-93%. The 1 H NMR spectrum of 8 displayed two meta-coupled protons at δH 7.88 and 8.18, three methoxy groups at δH 3.80 and 3.97 (x2), and one hydrogen-bond hydroxyl group at δH 15.71. NMR data of 8 and its mother compound 2 was highly similar. The differences are the disappearance of three aromatic protons (H-6, H-8, and H-5′), indicating that three positions were brominated. Altogether, the chemical structure of 8 was elucidated as shown. The identification of compounds 9-11 was readily established on the basis of their 1 H, 13 C NMR, and HRESI mass spectra (Tables S3 and S4).

Alpha-Glucosidase Inhibitory Activity of Isolated Compounds
The in vitro alpha-glucosidase inhibitory activity of 1-16 was evaluated (Table 1). Compounds 2, 4, 6-9, and 11-14 exhibited good inhibition of alpha-glucosidase with the IC 50 values were in the range of 30.5 to 282.0 µM, greater than the standard, acarbose (IC 50 332.5 µM). The C-3-substituted flavones 1-6 were relatively weaker than C-3-non-substituted analogues 7, 12, and 13, indicating the important role of the 3-H substituent in biological activity among flavones. As regards to the synthesized products, compounds 8-11 were more active than their parent compounds 2 and 3. The 6-brominated and 8-brominated products 9 and 10 were slightly stronger than their parent compound (3) whilst 11 significantly increased the activity (IC 50 30.5 µM). This indicated the activity preference for the 3 /5 -bromo positions. Similarly, 6,8,5 -tribromoayanin (8) was stronger than the starting material 2 (IC 50 87.1 µM). It is worth noting that the number of bromine atoms of the B-ring affected the biological activity.

Inhibition Type and Inhibition Constants of the Compounds 8 and 11 on Alpha-Glucosidase
In order to examine the inhibition mechanism of compound 11, their activity was measured at the different concentration of pNPG. The Lineweaver-Burk plots of a kinetic study of 8 and 11 showed linearity at each concentration examined (0, 4.96, 9.92, and 19.84 µM for 8 and 0, 21.62, 43.03, and 86.06 µM for 11), which all intersected the x-axis in the second quadrant ( Figure 2). The kinetic analysis revealed that V max decreased while Km remained constant, which showed that compounds 8 and 11 acted as noncompetitive inhibitors. The inhibition constants (Ki) of 8 was 39.82 µM and that of 11 was 198.87 µM.

Antibacterial Activity of Isolated Compounds
All compounds were evaluated for their in vitro antibacterial activity using the agar well diffusion method against methicillin-resistant Staphylococcus aureus (MRSA). Compounds 4 and 5 inhibited MRSA with diameters of inhibition zones of 14 and 15 mm, respectively, at the quantity of 50 µg for each compound, compared to that of the positive control, apramycin (24 mm). The others are inactive. All compounds failed to reveal any activity against ESBL-producing Escherichia coli.

In Silico Molecular Docking Model
The calculation results of molecular docking model of high active compound, 8, 11, and Acarbose based on autodock tools have been built in Figures 3-9, Figures S22-S27 and Table S5. In an in silico docking model, the most stable conformation of ligand 11 was bound to the active pocket on target enzyme 4J5T. Those interactions were assessed to be very strong because the values of Free Energy of Binding ∆G and the inhibition constant, K i between the most stable ligand 11 and target enzyme 4J5T have been calculated: −9.45 Kcal/mol and 0.118 µM, as shown in Figure 3 and Table S5. As indicated in Figure 4 and Table S5,  . Those hydrogen bonds were formed with hydrogen atoms of phenolic hydroxyl and bromine atoms of aromatic rings. For the hydrogen linked from the bromine atom of ligand to the oxygen atom of ILE362 of amino acid of A chain, the receptor was the strongest hydrogen, 1.83 Å among them. Those hydrogen bonds had hydrophilic interactions (green areas in Figure 3). As shown in Figure 5, the significant interactions were formed from active sites on the receptor to the most stable ligand such as classical hydrogen bonds; for instance, Asn 448 to bromo of phenyl ring, Arg 428 to oxygen of OH group of phenolic hydroxy, Val 446 to bromo atom of phenyl ring, Ile 362 to hydrogen atom of OH group of phenyl ring, and Gly 447 to hydrogen atom of phenolic hydroxy. Another non classical hydrogen bond was from Lys 363 to oxygen of hydroxy of phenolic hydroxy. The halogen interactions linked from Glu 429 and Gln 442 to bromine atoms of phenyl rings, respectively. The hydrophobic ligand interactions formed pi-sima from alkenyl group to Leu 364, pi-alkyl from Leu 364 to pi system of aromatic ring, and alkyl from Lys 363, Arg 428, and Phe 444 to the bromine atoms of aromatic rings. Other residual interactions were determined from Glu 361, Glu 443, and Gln 445, which linked weak interactions to the wall of cells. Briefly, the ligand 11 was the potential delivery drug because it identified well functional groups (via hydrogen bonds, halogen bonds), cap groups (via pi-sigma, alkyl, pi-alkyl interactions), and other Van Der Waals interactions, which bound weak to the wall of cells, as shown in Figure 5. Other the secondary interactions, which formed between ligand 11 and receptor 4J5T, were hydrogen bond, steric, and overlap interactions. Those interactions have established the stable interactions of the conformation ligand and receptor, as indicated in Figure 6. Those steric interactions (light blue), which linked from residual amino acids of A chain of 4J5T to active site atoms on 11 were Ile 362, Val 446, Glu 443, Phe 444, Arg 428, and Asn 448. The hydrogen bond (brown color) formed from Ile 362 and Gln 447 to the most stable conformation ligand 11. The steric interactions show violet circles on atoms of the conformation ligand. The bigger the effect of the steric, the bigger the sizes of the violet circles were, as shown in Figure 6. The interaction profile between the most stable ligand 8 and target enzyme is built in Figures 7-9, Table S5. As shown in Figure 7, the most stable conformation ligand 8 was selected to link to the target enzyme on enzyme pocket. Those interactions were assessed with the value of Free Energy of Binding ∆G and the inhibition constant K i , as shown in Figure 7, and calculated as −8.83 Kcal/mol and 0.337 µM. As indicated in Figure 8

Isolation and Structure Elucidation of the Compounds
Gravity column chromatography was performed on silica gel 60 (0.040-0.063 mm, Merck, Darmstadt, Germany). TLC for checking chromatographic patterns of fractions and isolated compounds was carried out on silica gel 60 F 254 (Merck) and spots were visualized by spraying with 10% H 2 SO 4 solution followed by heating. Specific rotations were obtained on a Jasco P-1010 polarimeter (Oklahoma City, OK, USA). The HR-ESI-MS were recorded on a MicroOTOF-Q mass spectrometer (Bruker, Billerica, MA, USA). The NMR spectra were measured on a Bruker Avance 500 MHz spectrometer (Bruker, Billerica, MA, USA).

General Procedure to Synthesize Compounds 8-11
In 4 mL of acetic acid, ayanin (2, 10 mg, 0.029 mmol) and sodium bromide (14.9 mg, 0.145 mmol) were dissolved at room temperature under stirring. The reaction was added to 0.02 mL (0.196 mmol) of 30% hydrogen peroxide. The reaction was conducted for 30 min and was periodically monitored by TLC. After neutralizing it with saturated sodium hydrogen carbonate, the mixture was extracted with ethyl acetate-water (1:1, v/v) to gain an organic layer. This was pooled, washed with brine, and dried over anhydrous Na 2 SO 4 . The residue was further absorbed onto column chromatography successively using the gradient system of n-hexane: chloroform (1:4, v/v) to afford 8 (14.9 mg). The procedure followed the previous report with modifications [26].
In 4 mL of acetic acid, kamatakenin (3, 10 mg, 0.032 mmol) and sodium bromide (16.4 mg, 0.160 mmol) were dissolved at room temperature under stirring. The reaction was added to 0.02 mL (0.196 mmol) of 30% hydrogen peroxide and was conducted in 30 min. The work-up followed the same procedure as mentioned previously to obtain the residue. The residue was further absorbed onto column chromatography successively using the gradient system of n-hexane: chloroform (1:4, v/v) to afford 11 (18.6 mg). The ratio of 3 and sodium bromide was modified to 1:1 and the reaction was repeated following the previously mention procedure. The reaction was conducted in 1 h and compounds 9 (4.4 mg) and 10 (6.0 mg) were obtained after silica gel CC.

Alpha-Glucosidase Inhibition Assay
The alpha-glucosidase (0.2 U/mL) and substrate (5.0 mM p-nitrophenyl-α-D-glucopyranoside) were dissolveed in 100 mM of pH 6.9 sodium phosphate buffer [27]. The inhibitor (50 µL) was preincubated with alpha-glucosidase at 37 • C for 20 min, and then the substrate (40 µL) was added to the reaction mixture. The enzymatic reaction was carried out at 37 • C for 20 min and stopped by adding 0.2 M of Na 2 CO 3 (130 µL). Enzymatic activity was quantified by measuring absorbance at 405 nm. All samples were analyzed in triplicate at five different concentrations around the IC 50 values, and the mean values were retained. The inhibition percentage (%) was calculated by the following equation: 3.6. Inhibitory Type Assay of 8 and 11 on Alpha-Glucosidase The mechanisms of inhibition of alpha-glucosidase by 8 and 11 were determined by Lineweaver-Burk plots (Microsoft Excel 2010, Washington, WA, USA), using methods similar to those reported in the literature [28]. Enzyme inhibition due to various concentrations of the 8 and

Antibacterial Activity Assay
Methicillin-resistant Staphylococcus aureus (MRSA), extended spectrum beta-lactamaseproducing Escherichia coli (ESBL-producing E. coli) and the agar well diffusion method were used to evaluate the antibacterial activity of the isolated compounds [29]. The strains were cultured in nutrient broth at 37 • C for 18 h and diluted with steriled 0.9% NaCl to obtain the 1.5 × 10 8 CFU/mL bacterial solution. Then, 100 µL of the bacterial solution was spread on Mueller-Hinton agar (MHA) plate on which 8-mm wells were created aseptically with tips. The isolated compounds were dissolved in dimethyl sulfoxide at the concentration of 1 mg/mL and 50 µL of each compound solution was applied in each well. The plates were incubated at 37 • C for 16-18 h and the diameters of the inhibtion zones were measured. Dimethyl sulfoxide and apramycin were used as controls in this experiment.

Molecular Docking Study Method
The calculations of the molecular docking model were performed according to the procedure in the supporting information file and followed the previous article [30].