Synthesis and Molecular Docking Studies of Alkoxy- and Imidazole-Substituted Xanthones as α-Amylase and α-Glucosidase Inhibitors

Current antidiabetic drugs have severe side effects, which may be minimized by new selective molecules that strongly inhibit α-glucosidase and weakly inhibit α-amylase. We have synthesized novel alkoxy-substituted xanthones and imidazole-substituted xanthones and have evaluated them for their in silico and in vitro α-glucosidase and α-amylase inhibition activity. Compounds 6c, 6e, and 9b promoted higher α-glucosidase inhibition (IC50 = 16.0, 12.8, and 4.0 µM, respectively) and lower α-amylase inhibition (IC50 = 76.7, 68.1, and >200 µM, respectively) compared to acarbose (IC50 = 306.7 µM for α-glucosidase and 20.0 µM for α-amylase). Contrarily, derivatives 10c and 10f showed higher α-amylase inhibition (IC50 = 5.4 and 8.7 µM, respectively) and lower α-glucosidase inhibition (IC50 = 232.7 and 145.2 µM, respectively). According to the structure–activity relationship, attaching 4-bromobutoxy or 4′-chlorophenylacetophenone moieties to the 2-hydroxy group of xanthone provides higher α-glucosidase inhibition and lower α-amylase inhibition. In silico studies suggest that these scaffolds are key in the activity and interaction of xanthone derivatives. Enzymatic kinetics studies showed that 6c, 9b, and 10c are mainly mixed inhibitors on α-glucosidase and α-amylase. In addition, drug prediction and ADMET studies support that compounds 6c, 9b, and 10c are candidates with antidiabetic potential.


Introduction
Diabetes mellitus is a metabolic disorder with increasing prevalence. This disease now afflicts about 537 million people worldwide according to the International Diabetes Federation, and thus represents a public health problem [1]. It is characterized by hyperglycemia stemming from congenital or acquired insulin secretion deficiency. The best therapeutic approach known to date consists of inhibiting intestinal enzymes responsible for carbohydrate hydrolysis, such as α-amylase and α glucosidase. Of the commercially available antidiabetic drugs, α-glucosidase inhibitors capable of acting in the intestine seem to be most effective for reducing postprandial hyperglycemia [2]. available antidiabetic drugs, α-glucosidase inhibitors capable of acting in the intestine seem to be most effective for reducing postprandial hyperglycemia [2]. α-Amylase is an enzyme (EC. 3.2.1.1) secreted by the pancreas and salivary glands that hydrolyzes α-linked polysaccharides (e.g., starch and glycogen) to maltose [3]. α-Glucosidase (EC. 3.2.1.20), an enzyme found in the small intestine, catalyzes the cleavage of the α-1,4-glycosidic bonds of maltose to form glucose [4]. Currently, antidiabetic α-amylase and α-glucosidase inhibitors (e.g., acarbose, miglitol, and voglibose) are based on carbohydrate-related structures and are effective. However, their use is limited by the adverse effects of flatulence, abdominal pain, and diarrhea, which could result from the fermentation of undigested carbohydrates derived from the strong inhibition of α-amylase [5][6][7]. Therefore, it is desirable to design new selective molecules with strong inhibition of α-glucosidase and weak inhibition of α-amylase to minimize side effects.
Several heterocyclic compounds containing oxygen and nitrogen are relevant for designing and developing new drugs. For instance, xanthone is a dibenzo-γ-pyrone heterocycle that has drawn the attention of researchers due to its broad spectrum of biological activity [8][9][10][11]. Xanthone derivatives are oxygenated heterocyclic compounds that occur as secondary metabolites in some families of higher plants (Guttiferae, Gentianaceae, Moraceae, Clusiaceae, and Polygalaceae). Prenylated xanthones such as α-mangostin ( Figure  1) exhibit a broad spectrum of biological activity, such as antimicrobial [12,13], antioxidant [14], antiviral [8], and anticancer properties [15,16], and also express multiple target proteins [17][18][19][20][21]. The functionalization of α-mangostin has led to antimicrobial and anticancer agents with significantly improved pharmacological properties [16,[22][23][24][25]. The functionalization of xanthones as α-glucosidase inhibitors has revealed the key factors involved in their observed inhibitory activity: the formation of an H-bond, The functionalization of xanthones as α-glucosidase inhibitors has revealed the key factors involved in their observed inhibitory activity: the formation of an H-bond, hy-drophobic groups with a π-conjugated system, and flexibility in conformation [21,26]. This is the case of 2-hydroxy-3-methoxyxanthone 1 (Figure 1), a natural product found in plants of the genus Hypericum (Clusiaceae), which exhibits anticancer properties [27][28][29]. Thus, it would be interesting to use this compound as a molecular platform to generate new derivatives with potential pharmaceutical applications.
Based on the aforementioned observations, the aim of the current contribution was to synthesize novel alkoxy-substituted xanthone derivatives and imidazole-substituted xanthones and test their potential as α-amylase and α-glucosidase inhibitors. Three structural elements were considered presently: (i) a natural xanthone core as an antidiabetic pharmacophore fragment, (ii) alkoxy groups substituted with a π-system or imidazolyl rings with drug-like properties, and (iii) a medium-chain alkoxy group to enhance lipophilicity. Compounds containing these elements were evaluated for their in vitro activity as α-amylase and α-glucosidase inhibitors. Moreover, in silico studies were performed to gain insight into the interaction of the compounds with the active site of α-amylase and α-glucosidase enzymes.

Chemistry
The synthetic pathways of alkoxy-substituted xanthone derivatives 5, 6a-d, 7, and 8 (Scheme 1) started from the acylation of phenol 2 with 2-iodobenzoyl chloride in the presence of boron trifluoride diethyl etherate, thus obtaining benzophenone 3a. The latter compound was O-alkylated by a reaction with methyl 2-bromoacetate (4a) to afford 3b, which was treated with N,N-dimethylformamide dimethyl acetal (DMFDMA) (4.0 mol equiv.) at 120 • C to provide the corresponding xanthonyl enaminone 5 and xanthone 6a in low to moderate yields. This pathway involves a cascade reaction that possibly occurs through intramolecular cyclization followed by condensation with DMFDMA. Xanthone 1 was obtained by the cyclization of the benzophenone intermediate 3a with a solution of KOH in water at 100 • C for 6 h [45]. Alkylation of compound 1 with α-halocarbonyls 4a-c or allyl bromide (4d) in the presence of K 2 CO 3 produced alkoxy-substituted xanthones 6a-d in high yields (Scheme 1). The oxyallyl xanthone 6d was subjected to a Claisen rearrangement to furnish allylhydroxy-substituted xanthone 7 in excellent yield. 2-Hydroxyxanthone 1 was also reacted with 1,1-diethoxy-3-methylbut-2-ene in the presence of 3-methylpicoline, leading to pyranoxanthone 8 in modest yield.
The two series of imidazole-substituted xanthones 10a-f and 12a-f were prepared by attaching imidazoles to the xanthones 9a-b and 11a-b, respectively. Firstly, alkylation of the phenoxy group of xanthones 1 and 7 with 1,4-dibromobutane under base conditions generated xanthones 9a-b in excellent yields. Substitution of the bromine atom by the 2-substituted imidazoles 13a-c resulted in the corresponding imidazole-substituted xanthones 10a-f in moderate to good yields [22].
For 12a-b, the intermediate epoxides 11a-b were formed by alkylation of 1 and 7 with epichlorohydrin in ethanol in the presence of KOH. The epoxide ring of derivatives 11a-b was opened with imidazoles 13a-c in methanol to convert them into the respective compounds 12a-f in moderate to good yields [25]. The alkylation of 7 with α-halocarbonyl 4c delivered alkoxy-substituted xanthone 6e in excellent yield (Scheme 2). The two series of imidazole-substituted xanthones 10a-f and 12a-f were prepared by attaching imidazoles to the xanthones 9a-b and 11a-b, respectively. Firstly, alkylation of the phenoxy group of xanthones 1 and 7 with 1,4-dibromobutane under base conditions generated xanthones 9a-b in excellent yields. Substitution of the bromine atom by the 2substituted imidazoles 13a-c resulted in the corresponding imidazole-substituted xanthones 10a-f in moderate to good yields [22].
For 12a-b, the intermediate epoxides 11a-b were formed by alkylation of 1 and 7 with epichlorohydrin in ethanol in the presence of KOH. The epoxide ring of derivatives 11a-b was opened with imidazoles 13a-c in methanol to convert them into the respective compounds 12a-f in moderate to good yields [25]. The alkylation of 7 with α-halocarbonyl 4c delivered alkoxy-substituted xanthone 6e in excellent yield (Scheme 2). All the synthesized compounds were fully characterized by 1 H-NMR, 13 C-NMR, and HRMS. In the 1 H-NMR spectra, for compounds 6a-e, the methylene protons of the acetophenone moiety appeared as a singlet at 4.74-5.74 ppm. For compounds 10a-f, the methylene protons adjacent to the nitrogen atom of the butoxyimidazole moiety were observed as triplets at 4.03-4.22 ppm, while compounds 12a-f were observed as a doublet of doublets at 4.06-4.22 ppm. All the peak values from 6.84-8.32 ppm were assigned to aromatic protons. In the 13 C-NMR spectra, the carbonyl groups of xanthone and acetophenone appeared at 174.5-175.9 and 193.3-203.7 ppm, respectively.
In the case of xanthonyl enaminone 5, a single stereoisomer was obtained, and its Z geometry was established by NOE experiments. Irradiation of the signal assigned to the methyl protons of the dimethylamine group produced the enhancement of the signal corresponding to the aryloxy ring of the xanthone scaffold. This stereoselectivity has been observed in similar systems, probably because of the greater stability of the planar πconjugated acrylate system when the bulky dimethylamine group is located at the opposite side of the double bond [46]. All the synthesized compounds were fully characterized by 1 H-NMR, 13 C-NMR, and HRMS. In the 1 H-NMR spectra, for compounds 6a-e, the methylene protons of the aceto phenone moiety appeared as a singlet at 4.74-5.74 ppm. For compounds 10a-f, the meth ylene protons adjacent to the nitrogen atom of the butoxyimidazole moiety were observed as triplets at 4.03-4.22 ppm, while compounds 12a-f were observed as a doublet of dou blets at 4.06-4.22 ppm. All the peak values from 6.84-8.32 ppm were assigned to aromatic protons. In the 13 C-NMR spectra, the carbonyl groups of xanthone and acetophenone ap peared at 174.5-175.9 and 193.3-203.7 ppm, respectively.
In the case of xanthonyl enaminone 5, a single stereoisomer was obtained, and its Z geometry was established by NOE experiments. Irradiation of the signal assigned to the methyl protons of the dimethylamine group produced the enhancement of the signal cor responding to the aryloxy ring of the xanthone scaffold. This stereoselectivity has been observed in similar systems, probably because of the greater stability of the planar π-con jugated acrylate system when the bulky dimethylamine group is located at the opposite side of the double bond [46].

In Vitro α-Glucosidase Inhibition
After testing each compound for its inhibitory effect on α-glucosidase, this result was compared to the effect of acarbose (14), an antidiabetic drug known to inhibit the α-glucosidase and α-amylase enzymes [6]. Initially, the compounds were evaluated at 400 µM. Structurally, alkoxy-substituted xanthones were divided into two groups based on the nature of the alkoxy chain substituents at the C-1 and/or C-2 positions of the xanthone core: (1) substituted alkoxy derivatives 5, 6a-e, 8, 9a-b, and 11a-b and (2) substituted imidazolyl derivatives 10a-f and 12a-f. Xanthone 1 at 400 µM exhibited weak inhibitory activity on α-glucosidase (Table 1), as did 6a (with a 2-oxyacetate substituent at the 2hydroxy group). An insertion of the (4-chlorophenyl)-2-oxoethoxy substituent afforded 6c, increasing the inhibitory effect (IC 50 = 143.6 ± 0.17 µM for 6a and IC 50 = 16.0 ± 0.03 µM for 6c). Contrarily, the presence of an α-acetonyl or allyl group (6b and 6d) resulted in a weak inhibition effect. Similarly, a significant decrease in inhibitory activity was found with 5, formed by the attachment of an enaminone moiety to the 2-oxyacetate group of 6a. Interestingly, 7 (with an allyl group at the C-1 position) generated a greater inhibitory effect on α-glucosidase (IC 50 = 196.4 ± 0.07 µM) than its analogs 1 and 8. The alkoxy-substituted xanthone derivatives 6e, 9a-b, and 11a-b were examined to clarify the role of the allyl group, which in 9b provided the most potent inhibitory activity (IC 50 = 4.00 ± 0.007 µM). The introduction of an imidazole ring instead of the bromine group at the C-4 position of the chain in 9b furnished compounds 10a-d and 10f, resulting in a significant decrease in the inhibitory effect. Regarding the imidazolyl-substituted xanthones series 12a-f, 12b and 12e (with a phenyl group in the 2-imidazole ring) showed higher inhibitory activity (IC 50 = 112.8 ± 0.12 µM and 104.9 ± 0.01 µM, respectively) than 12a and 12d. According to the results, the introduction of the imidazolyl group had no significant effect on activity since almost all the products (10a-d, 10f, and 12a-e) are less effective than their intermediates (9a-b and 10a-b).
In summary, the inhibition of α-glucosidase produced by acarbose (14, the reference drug) was significantly improved when a phenyl or aryl ring was present at the C-2 side chain combined with the 2-substituted imidazole ring of the xanthone core, leading to 6c, 6e, 9a-b, 12b, and 12f. Considering structural similarities among other molecules, inhibition was much stronger for those containing the C-1 allyl group (7, 9b, and 11b). The data suggest that a molecule with an allyl or substituted aryl group has better affinity for the amino acid residues of the α-glucosidase enzyme and thus enhanced bioactivity, which owes itself to π-sticking or hydrophobic effects [11,21,37], as well as interactions with the halogen atoms substituted at the phenyl groups or at the alkyl side chain [36,38,47].

In Vitro α-Amylase Inhibition
All compounds evaluated on α-glucosidase were tested for their capacity to diminish α-amylase activity ( Table 1). The compounds were evaluated at 100 µM. Natural xanthone 1 without any substituent modification at the C-2 position exerted very strong inhibition of α-amylase. The incorporation of a 2-oxyacetate group at C-2 in compound 6a did not show activity. A limited inhibitory effect was observed with the insertion of a (4-chlorophenyl)-2oxoethoxy substituent in compound 6c or a 4-bromobutoxy substituent in compound 9a. The IC 50 value was higher for 6c and 9a than for acarbose (14), indicating a lower inhibitory effect for these two compounds. No inhibitory activity was detected for 7 with an allyl group at the C-1 position, while a small inhibitory effect was found for 6e, which was prepared by introducing a (4-chlorophenyl)-2-oxoethoxy substituent at C-2. There was limited inhibitory activity with 9a and sharply increased activity when the bromine atom at the side chain of this analogue was replaced by a 2-(4-chlorophenyl)imidazol-1-yl group to provide 10c.
This increase also occurred for 10f. On the other hand, the introduction of an additional OH group at the alkoxy side chain of compounds 12a-f led to an absence of inhibitory activity.
In summary, greater inhibition of α-glucosidase and lesser inhibition of α-amylase were achieved by the incorporation of a (4-chlorophenyl)-2-oxoethoxy moiety (6c and 6e) and 4-bromobutoxy (9b) at C-2 of the xanthone scaffold. In contrast, the lowest inhibition of α-glucosidase and highest inhibition of α-amylase were found with natural xanthone 1 and with the addition of the 4-(2-(4-chlorophenyl)imidazol-1-yl)butoxy moiety (i.e., 10c and 10f). Considering the aforementioned desirability of a low inhibitory effect on αamylase (to avoid gastrointestinal side effects) together with potent inhibitory activity on α-glucosidase, compounds 6c, 6e, and 9b are promising candidates for the development of antidiabetic drugs.

Enzymatic Kinetic Study
In order to explore the interaction mechanism of alkoxy-xanthones 6c, 9b, and 10c, the type of inhibition exhibited by these selective inhibitors was analyzed using Lineweaver-Burk plots (double reciprocal). The X-axis values represent the reciprocal for the αglucosidase substrate, p-nitrophenyl-α-D -glucopyranoside (p-NPG), while for the α-amylase substrate they represent starch, thus being 1/(starch). The Y-axis values are the reciprocal of the reaction velocity (Vo), thus being 1/Vo. Given that the plots did not intersect the Xor Y-axis, the inhibition of α-glucosidase exerted by these compounds is carried out in mixed mode ( Figure 2A). The K I values of 6c, 9b, and 10c are 21.5, 1.25, and 139.0 µM, respectively. The K I values for these compounds are less than K m , indicating that they have a higher affinity for the enzyme than the substrate used in the assay [48].

Evaluation of Antioxidant Activity
Since antioxidants contribute to the prevention of diabetes mellitus and other diseases [49], the antioxidant potential of the synthesized compounds was determined. The capacity for free radical scavenging was assessed by means of the 2,2-diphenyl-1picrylhydrazyl (DPPH) radical method, with butylhydroxytoluene (BHT) as the positive control. A decrease in color intensity represents the scavenging of DPPH (Table 1), which was calculated as a percentage. The alkoxy-substituted xanthones 6a-e, 9a-b, and 11a-b, as well as imidazole-substituted xanthones 10a-f and 12a-f, did not show any significant antiradical activity even at the maximum concentration tested. Only compounds 1 and 7 were able to scavenge DPPH to some extent (ca. 33% and 46% at 2.5 mM), suggesting that this effect is mainly related to the hydroxy group substituted at the C-2 position. The amylase plots made it possible to determine that compound 6c is a competitive type of inhibitor, while 10c is a mixed-type inhibitor. The KI values of 6c and 10c are 63.2 and 2.2 µM, respectively ( Figure 2B). These KI values indicate that 6c has greater affinity for the enzyme.

Evaluation of Antioxidant Activity
Since antioxidants contribute to the prevention of diabetes mellitus and other diseases [49], the antioxidant potential of the synthesized compounds was determined. The capacity for free radical scavenging was assessed by means of the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical method, with butylhydroxytoluene (BHT) as the positive control. A decrease in color intensity represents the scavenging of DPPH (Table 1), which was calculated as a percentage. The alkoxy-substituted xanthones 6a-e, 9a-b, and 11a-b, as well as imidazole-substituted xanthones 10a-f and 12a-f, did not show any significant antiradical activity even at the maximum concentration tested. Only compounds 1 and 7 were able to scavenge DPPH to some extent (ca. 33% and 46% at 2.5 mM), suggesting that this effect is mainly related to the hydroxy group substituted at the C-2 position. The amylase plots made it possible to determine that compound 6c is a competitive type of inhibitor, while 10c is a mixed-type inhibitor. The K I values of 6c and 10c are 63.2 and 2.2 µM, respectively ( Figure 2B). These K I values indicate that 6c has greater affinity for the enzyme.

Molecular Docking Analysis
To explore the binding interactions of the most active compounds, molecular docking studies were carried out between alkoxy-substituted xanthones 6c, 6e, 9b, and 10c and the isomaltase enzyme (the α-glucosidase of S. cerevisiae), as well as 1, 6c, 9a, 10c, and 10f and the human α-amylase enzyme. The results of the interactions are illustrated in 2D and 3D (Figures 3 and 4), revealing that the alkoxy-substituted xanthone derivatives recognized some of the key amino acid residues in the catalytic pocket, such as His112, Arg213, Asp215, Glu277, His351, Asp352, and Arg442. A similar set of residues is reported by maltose inhibitors [50][51][52][53][54].

Molecular Docking Analysis
To explore the binding interactions of the most active compounds, molecular docking studies were carried out between alkoxy-substituted xanthones 6c, 6e, 9b, and 10c and the isomaltase enzyme (the α-glucosidase of S. cerevisiae), as well as 1, 6c, 9a, 10c, and 10f and the human α-amylase enzyme. The results of the interactions are illustrated in 2D and 3D (Figures 3 and 4), revealing that the alkoxy-substituted xanthone derivatives recognized some of the key amino acid residues in the catalytic pocket, such as His112, Arg213, Asp215, Glu277, His351, Asp352, and Arg442. A similar set of residues is reported by maltose inhibitors [50][51][52][53][54].
Acarbose (14) Figure 3. Representation of the interactions of alkoxy-substituted xanthones 6c, 6e, and 9b within the active pocket of isomaltase. The 3D models illustrate the interactions with the amino acid residues of the catalytic pocket of the enzyme. In the 2D model, conventional hydrogen (dark green dotted lines) and carbon hydrogen bonds (light green) are portrayed, as well as π-sigma (purple), π-π T-shaped and π-π stacked (fuchsia), π-alkyl and alkyl (pink), and π-anion and π-cation (orange) interactions. The amino acids are depicted with circles of different colors (pink (basic), orange (acid), cyan (polar), and yellow (non-polar)). Representation of the interactions of alkoxy-substituted xanthones 6c, 6e, and 9b within the active pocket of isomaltase. The 3D models illustrate the interactions with the amino acid residues of the catalytic pocket of the enzyme. In the 2D model, conventional hydrogen (dark green dotted lines) and carbon hydrogen bonds (light green) are portrayed, as well as π-sigma (purple), π-π T-shaped and π-π stacked (fuchsia), π-alkyl and alkyl (pink), and π-anion and π-cation (orange) interactions. The amino acids are depicted with circles of different colors (pink (basic), orange (acid), cyan (polar), and yellow (non-polar)).
Analysis of the docking data for natural xanthone 1 showed hydrogen bond interactions with carboxylate groups of the amino acid of the catalytic triad (Asp197, Glu233, and Asp300), suggesting competitive-type inhibition. Regarding compounds 10c and 10f, the dibenzo-γ-pyrone system of xanthone displays hydrophobic π-π and π-alkyl interactions with residues Trp59 and Tyr62. The fragment 2-(4-chlorophenyl)butoxy imidazole at C-2 of the xanthone core is involved in hydrophobic interactions, such as π-π and π-alkyl (Tyr62, Leu162, His 201, and Ala198), alkyl (Leu162, Ile235, Lys200), and π-anion (Asp300) interactions. In addition, unconventional hydrogen bonding interactions are observed with Asp197 and Asp300 residues of the catalytic triad. This could be due to the fact that compounds 10c (Table S1) and 10f adopted U-shaped and V-shaped conformations, respectively, modifying their rearrangement in space and generating these interactions close to the catalytic pocket, thus suggesting competitive-type inhibition [55]. Data of 6c and 9a are summarized in Table S2 and Figure S90, Supplementary Materials.

Prediction of Drug-like Properties
The physicochemical properties of the synthesized compounds were analyzed using the OSIRIS DataWarrior program [59] and are summarized in Table S3 (Supplementary  Materials). Almost all compounds (except 10f and 12f) have a molecular weight under 500 g/mol. Lipophilicity is expressed as log p and represents the affinity of a molecule or a moiety for a lipophilic environment. Values close to 5 indicate high permeability of lipids, while negative values evidence low permeability [60].
Compounds 1, 6a, 7, 9a, 11b, and 12b-e showed acceptable log p values within the range of 2.78-4.96. Slight to moderate solubility in water was found for all compounds, with log S values ranging from −4.64 to −7.84 [61]. Acarbose (14) has a value of 0.58 (its polar groups form hydrogen bonds with water). The polar surface area (PSA) of a molecule is the sum of the surfaces of oxygen or nitrogen atoms and the hydrogen atoms attached to them. For a drug to cross the blood-brain barrier, the PSA value must be less than 90 A 2 [62]. All derivatives herein evaluated met this requirement. Since 1, 6a, 6c, 7, 9a, 10c,  11b, and 12b-e comply with Lipinski's rule of five (Table S3), they are expected to have oral bioavailability.

Prediction of Druglikeness, ADME Properties, and Toxicity
The most potent α-glucosidase and α-amylase inhibitors (6c, 6e, 9a, 10c, and 10f) were analyzed with the online software PreADMET 2.0 to predict their druglikeness, ADME, and toxicity (Table 4) [63]. All compounds except 10f complied with Lipinski's rule of five. Regarding the intestinal barrier, represented by Caco-2 cells, the permeability of 6c, 6e, 10c, and 10f was moderate, while that of 9a was poor. All five compounds have good human intestinal absorption (HIA) and acceptable permeability of the blood-brain barrier (BBB) and skin . Compounds 6e, 10c, and 10e are non-mutagenic, and 6c, 6e, 10c, and 10f are non-carcinogenic on mice and rats. Compound 9a had a carcinogenic effect on rats but not on mice. Furthermore, there was a medium risk of cardiotoxicity (hERG inhibition) for all five compounds when examined in silico. It is worth mentioning that the biological activity data, together with in silico studies (docking, drug prediction, and ADMET), allow for a broader perspective of the effects that the 4 -chlorophenylacetophenone and 4-bromobutoxy moieties produce on the glucosidase and amylase targets, suggesting structural features for the design of new molecules with antidiabetic properties.

α-Amylase Inhibition Assay
α-Amylase inhibitory activity was quantified according to the method developed by Chokki et al., with some modifications [65]. The reaction mixture consisting of 50 µL of 0.1 M phosphate buffer (pH 6.8), 10 µL of α-amylase solution (5.0 unit/mL), and 20 µL of the sample at various concentrations (from 100 µM to 5.0 µM) was placed in a 96-well plate and pre-incubated at 37 • C for 15 min, and 20 µL of 1% soluble starch (0.1 M phosphate buffer, pH 6.8) was then added as a substrate and incubated at 37 • C for 45 min. Finally, 100 µL of 3,5-dinitrosalicylic acid (DNS) was added and heated at 100 • C for 20 min, and absorbance was then read at 540 nm in a microplate reader (Epoch, BioTek ® ). The reaction system without any test compound was used as the control, and the system without α-amylase served as a blank for correcting background absorbance. The percentage of α-amylase inhibition was calculated for each sample with Equation (2): All measurements were performed in quadruplicate and the values are expressed as the mean ± standard deviation.

Kinetic Study
To explore the type of enzyme inhibition, kinetic studies were carried out with αglucosidase and α-amylase using a methodology such as that described in the inhibitory activity assays. The alkoxy-substituted xanthones were evaluated at four different concentrations according to their IC 50 . Various concentrations of substrates were used for each of the enzymes in the range of 0.5-5.0 mM for p-NPG in α-glucosidase and 0.1-1.0% for α-amylase. The type of inhibition for each test compound was determined by utilizing double reciprocal plots. Inhibition constants (K I ) were calculated from substrate versus reaction rate curves using nonlinear regression of the enzyme inhibition kinetic function [55,65].

DPPH Radical Scavenging Assay
The scavenging of free radicals by the synthesized compounds was assessed based on the previously reported DPPH radical assay, with slight modifications [66]. The reaction mixture consisted of 50 µL of compound in DMSO at various concentrations (from 2.5 mM to 0.2 mM) and 150 µL of DPPH solution at 133.33 µM in absolute ethanol. The reaction components were added at a ratio of 1:3 (v/v). The mixture was incubated at 37 • C for 30 min before absorbance was read at 517 nm using a microplate reader (Epoch, BioTek ® ). Butylhydroxytoulene (BHT) served as the positive control. Scavenging capacity (%) is expressed as the percentage decrease in DPPH.

SC% = [(A control − A test )/A control ] × 100
where A control is absorbance of the DPPH solution (control) and A test is the absorbance of the solution of DPPH and one of the compounds.

Docking Studies
The molecular docking studies were carried out in the AutoDock 4 program [67] using the crystallized proteins of isomaltase from Saccharomyces cerevisiae (PDB: 3A4A) and human pancreatic α-amylase (PDB: 1B2Y) in complex with the inhibitor acarbose. In these proteins, water molecules were removed, hydrogen atoms were added to the polar atoms (considering pH at 7.4), and Kollman charges were assigned with AutoDock Tools 1.5.6. The 3D structures of acarbose (14), 1 (natural xanthone), alkoxy-xanthones 6c, 6e, 9a, and 9b, and imidazole-xanthones 10c and 10f were sketched in two dimensions (2D) with ChemSketch and then converted to 3D in a mol2 format using the Open Babel GUI program [68]. The ligands were optimized with PM6 on Gaussian 98 software to obtain the lowest energy conformation. All the possible rotatable bonds, torsion angles, atomic partial charges, and non-polar hydrogens were determined for each ligand. In AutoDockTools, the grid dimensions for α-glucosidase were 78 × 60 × 78 Å 3 with points separated by 0.375 Å and centered at X = 26.313, Y = −3.544, and Z = 26.146. The grid dimensions for α-amylase were 90 × 70 × 66 Å 3 with points separated by 0.375 Å and centered at X = 16.758, Y = 8.692, and Z = 49.959. The hybrid Lamarckian genetic algorithm was applied for minimization and utilized default parameters. A total of one hundred docking runs were conducted to determine the conformation with the lowest binding energy (kcal/mol), which was adopted for all further simulations. AutoDockTools was used to prepare the script and files as well as to visualize the docking results, and these were edited with the Discovery 4.0 client.

Conclusions
Alkoxy-and imidazole-substituted xanthones 6-12 were synthesized and their inhibitory activity on α-glucosidase and α-amylase enzymes was evaluated. Compared to the reference drug acarbose (14), the inhibitory activity of 6c, 6e, and 9b was higher for α-glucosidase and lower for α-amylase, reflecting a desirable outcome. Based on structureactivity analysis of the results, a 4-bromobutoxy or 4 -chlorophenylacetophenone moiety in the molecule favors greater inhibition of α-glucosidase versus α-amylase. In contrast, inserting a 2-(4-chlorophenyl)butoxyimidazole moiety (10c) produces lower α-glucosidase inhibition and higher α-amylase inhibition. The mechanism of the enzymatic inhibition of 6c, 10c, and 9b was determined, establishing that for α-glucosidase they are mixed inhibitors, while for α-amylase, 6c is a competitive inhibitor and 10c is mixed. The docking studies revealed that the π-stacking and hydrophobic effects of the aromatic moiety at the C-2 position of the xanthone backbone play a key role in the interaction with the active sites of both α-glucosidase and α-amylase. Additionally, drug prediction and ADMET studies suggest that compounds 6c, 6e, and 9b are candidates for the development of new selective α-glucosidase inhibitors with antidiabetic potential.