Synthesis and α-Glucosidase Inhibition Activity of 2-[3-(Benzoyl/4-bromobenzoyl)-4-hydroxy-1,1-dioxido-2H-benzo[e][1,2]thiazin-2-yl]-N-arylacetamides: An In Silico and Biochemical Approach

Diabetes mellitus (DM) is a chronic disorder and has affected a large number of people worldwide. Insufficient insulin production causes an increase in blood glucose level that results in DM. To lower the blood glucose level, various drugs are employed that block the activity of the α-glucosidase enzyme, which is considered responsible for the breakdown of polysaccharides into monosaccharides leading to an increase in the intestinal blood glucose level. We have synthesized novel 2-(3-(benzoyl/4-bromobenzoyl)-4-hydroxy-1,1-dioxido-2H-benzo[e][1,2]thiazin-2-yl)-N-arylacetamides and have screened them for their in silico and in vitro α-glucosidase inhibition activity. The derivatives 11c, 12a, 12d, 12e, and 12g emerged as potent inhibitors of the α-glucosidase enzyme. These compounds exhibited good docking scores and excellent binding interactions with the selected residues (Asp203, Asp542, Asp327, His600, Arg526) during in silico screening. Similarly, these compounds also showed good in vitro α-glucosidase inhibitions with IC50 values of 30.65, 18.25, 20.76, 35.14, and 24.24 μM, respectively, which were better than the standard drug, acarbose (IC50 = 58.8 μM). Furthermore, a good agreement was observed between in silico and in vitro modes of study.


Introduction
Diabetes mellitus (DM) is an endocrine system illness and causes metabolic ailment that leads to hyperglycemia. DM is characterized by increased blood sugar levels either due to the body's reduced production of insulin or ineffective utilization of available insulin by the body. Chronic hyperglycemia results in neuron oxidative damage while

Introduction
Diabetes mellitus (DM) is an endocrine system illness and causes metabolic ailment that leads to hyperglycemia. DM is characterized by increased blood sugar levels either due to the body's reduced production of insulin or ineffective utilization of available insulin by the body. Chronic hyperglycemia results in neuron oxidative damage while prolonged hyperglycemia causes the production of lower caloric content of sugar in different organs [1,2]. Although genetic predisposition can be a factor in the development of DM, other major causing factors may include sedentariness, smoking, and intake of high-saturated fatty meals [3,4]. Change in lifestyle can decrease the risk of type 2 diabetes by 58%, for example, overweight people can avoid it by losing weight, a fact explained experimentally by the Finnish prevention trial [5] and the US diabetes prevention program [6]. DM also enhances the chances of cardiovascular diseases [7]. People suffering with hypertension are already leading a risky life due to high blood pressure [8] but they often develop DM and this combination of risks increases cardiovascular transience. The rate of DM is growing rapidly worldwide especially in developing countries [9]. The International Diabetes Federation (IDF) has observed that 371 million humans are infected with T2DM (type 2 diabetes mellitus) while an increase rate of 2% annually has been estimated. α-Amylase and α-glucosidase are the two enzymes that have a key role in the breakdown of polysaccharides to monosaccharides [10]. α-Glucosidase enzyme, released from mucosal cells, usually converts polysaccharides into monosaccharides causing a gradual increase in the absorption of glucose and hence increases intestinal blood glucose level [11]. These enzymes have role in breakdown and absorption of polysaccharides, so by inhibiting the activity of these enzymes, glucose levels in diabetic patients can be controlled even after consumption of carbohydrate-rich diets [12].
.  Our research group is involved in the search for novel and more effective antidiabetic agents [41][42][43][44][45], and in this context we have synthesized novel 1,2-benzothiazine-N-arylacetamides and screened them for their in silico and in vitro α-glucosidase inhibitory potential.
Our research group is involved in the search for novel and more effective antidiabetic agents [41][42][43][44][45], and in this context we have synthesized novel 1,2-benzothiazine-Narylacetamides and screened them for their in silico and in vitro α-glucosidase inhibitory potential.

Spectroscopic Characterization
All the synthesized derivatives were well characterized with 1 H-NMR, 13 C-NMR, and MS spectroscopic techniques. In the 1 H-NMR spectra, the methylene (CH2) protons appeared as a singlet at 3.95-4.20 ppm and the observation was consistent with literature values of similar compounds [46,49]. The methyl protons (CH3) and methoxy protons appeared as singlets at 1.96-2.17 and 3.64-3.86 ppm, respectively. The NH and enolic-OH protons were observed as singlets at 9.16-10.54 and 14.92-15.18 ppm, respectively [37,49,50]. Peaks for enolic-OH protons were not observed in some spectra due to the exchangeable nature of OH protons [49]. All the peak values from 6.50-8.50 ppm were Scheme 1. Synthesis of 1,2-benzothiazine-N-arylacetamides 11a-m and 12a-m.

Spectroscopic Characterization
All the synthesized derivatives were well characterized with 1 H-NMR, 13 C-NMR, and MS spectroscopic techniques. In the 1 H-NMR spectra, the methylene (CH 2 ) protons appeared as a singlet at 3.95-4.20 ppm and the observation was consistent with literature values of similar compounds [46,49]. The methyl protons (CH 3 ) and methoxy protons appeared as singlets at 1.96-2.17 and 3.64-3.86 ppm, respectively. The NH and enolic-OH protons were observed as singlets at 9.16-10.54 and 14.92-15.18 ppm, respectively [37,49,50]. Peaks for enolic-OH protons were not observed in some spectra due to the exchangeable nature of OH protons [49]. All the peak values from 6.50-8.50 ppm were assigned to aromatic protons. In MS spectra, the observed molecular ion peaks were found in accordance with the calculated values.

Molecular Docking (In Silico Analysis)
In silico screening is an excellent approach for the screening of libraries of compounds in a short time and thus it minimizes the laborious synthetic work [52][53][54][55]. Molecular docking is a widely used method to predict binding interactions between the 3D conformations of various ligands and receptor proteins that helps in optimization and leads toward drug development [56][57][58].
Regarding the effectiveness of this tool, we executed molecular docking studies of all the synthesized 1,2-benzothiazine derivatives. Molecular docking was performed via MOE software to screen out the compounds having the best residue binding interactions with the receptor protein (α-glucosidase). These compounds were ranked on the basis of root-mean-square deviation (rmsd) values, bindings with the selected residues (Asp203, Asp542, Asp327, His600, Arg526) ( Figure S1), and docking scores [59].
In the molecular docking experiment, most of the compounds showed good binding scores and binding interactions with the selected residues (Asp203, Asp542, Asp327, His600, Arg526) (Table 1, Figure 2). In silico studies of salacinol having a cyclic thiosugar sulfonium and a sulfate group in its structure and its derivatives found it blocked the α-glucosidase activity via bindings with the Asp203, Asp327, and Asp542 residues [60]. Recently, in silico analysis of anthocyanidins and anthocyanins was carried out that showed these scaffolds to be a good inhibitor of α-glucosidase enzymes. These compounds also showed their inhibitory action by interacting with the selected residues, Asp203, Asp327, and Asp542 [61]. Similarly, aglycone of curculigoside A and its derivatives also exhibited good in silico inhibition of α-glucosidase enzymes via interactions with Asp203, Asp327, and Asp542 residues [62]. The ligands 12a, 12d, 12e, and 12g having docking scores from −13.47 to −14.23 Kcal/mol were ranked the most effective antidiabetic scaffolds. Moreover, compounds 11c, 11h, 11i, and 12m were also considered important due to good binding energy values in the range of −12.13 to −12.80 Kcal/mol. These above-described ligands showed binding energy values close to the standard drug, acarbose (−16.18 Kcal/mol) and were tightly packed in the selected pocket of the receptor enzyme as shown in 3D maps ( Figure 3). The 2D interaction modes revealed that most of the ligands bind with Asp203, Asp327, and Asp542 among the selected residues and were supposed to be responsible for the α-glucosidase inhibitory activity ( Figure 2, Table S1). Acarbose (control) also showed good interactions with the selected residues, His600, Asp542, Arg526, and Asp327, in the As a whole, 1,2-Benzothiazine-N-arylacetamides 11a-m were observed to be less effective compared to derivatives 12a-m regarding their in silico α-glucosidase inhibitory potential (Table 1). However, compounds 11c, 11h, and 11i showed comparatively better in silico enzyme inhibitions as revealed from their low binding energies and rmsd values 11c 12a 12d 12e 12g Acarbose As a whole, 1,2-Benzothiazine-N-arylacetamides 11a-m were observed to be less effective compared to derivatives 12a-m regarding their in silico α-glucosidase inhibitory potential (Table 1). However, compounds 11c, 11h, and 11i showed comparatively better in silico enzyme inhibitions as revealed from their low binding energies and rmsd values ( Table 1). The compound 11c was observed as the most effective derivative in terms of binding energy and rmsd value, which were found to be −12.80 Kcal/mol and 1.08 Å, respectively. This ligand showed interactions with Asp327, Asp474, and Met444 residues. Asp327 interacted with the benzene ring of the benzothiazine nucleus. Asp474 formed a hydrogen bond with the chloro and Met444 with the oxygen atom of the SO 2 group. The ligand 11h exhibited a good docking score (−12.13 Kcal/mol) and low rmsd value (1.21 Å). A hydrogen bonding was observed between the NH group and the Asp542 residue while Thr205 interacted with the oxygen of the CO group. Similarly, a low docking score (−12.25 Kcal/mol) and rmsd value (1.11 Å) were found for the compound 11i. This is explained by the strong hydrogen bonding of this compound with the selected residue, Asp542, via NH functionality.
1,2-Benzothiazine-N-arylacetamides 12a-m preferentially exhibited good binding energies and better interactions with the selected residues compared to 1,2-benzothiazine-N-arylacetamides 11a-m (Table 1). Compound 12a was observed as the most potent compound as it showed a low docking score (−14.23 Kcal/mol), small rmsd value (1.21 Å), and good interactions with selected residues (Arg526, Asp542, Asp327). This ligand exhibited a hydrogen bond with Arg526 via the oxygen atom of the OH group. A second hydrogen interaction was found between the nitrogen atom of the NH group and the Asp542 residue. Similarly, Asp327 interacted with the benzene ring of the benzothiazine nucleus of ligand 12a. In addition, the present ligand also formed an interaction with Trp406 using the oxygen atom of the SO 2 group. These interactions and smaller rmsd value are responsible for the low binding energy value of ligand 12a. The compounds 12d, 12e, and 12g also showed good docking scores (−13.91, −13.47, and −13.52 Kcal/mol, respectively) due to dipole-dipole interactions with Asp203 and Asp327 via their OH and Br functionalities, respectively ( Figure 2). These ligands (12d, 12e, and 12g) also exhibited low rmsd values of 1.13, 1.27, and 1.01 Å, respectively. In the ligand 12m, the bromo group showed an interaction with the selected residue, Asp327. Due to this interaction and a low rmsd value of 1.19 Å, ligand 12m showed a good binding energy value of −12.32 Kcal/mol.
The compound 12k was found somewhat less effective as an inhibitor as revealed from its docking score and rmsd value ( Table 1). The compound 12k showed interactions with Asp327 and Tyr605 via both the bromo groups individually. The other members of this series showed high rmsd values and no interactions with the selected residues. In addition, the 1,2-benzothiazines 9a,b were also docked into the receptor pocket but these proved to be noninhibitors of the α-glucosidase enzyme (Table 1 and Table S1).
It is concluded from the above discussion that not only the benzothiazine ring system but also the presence of the acetamide group was found indispensable for the in silico inhibition of α-glucosidase enzymes. Furthermore, it was noticed that the presence of polar as well as nonpolar groups in the structures of scaffolds played a crucial role toward the inhibition of enzymes by interacting with the selected residue. It is clear from the 2D diagrams of molecules that the interactions with the selected residues groups, such as CO, CONH, NH, SO 2 , Cl, Br, OH, phenyl, and aryl, all participated equally. Compounds with low rmsd values having good binding energies and interactions with the selected residues were of our interest.

α-Glucosidase Inhibition (In Vitro Analysis)
In order to assess the in silico results, all the compounds were also subjected to the in vitro inhibition of α-glucosidase enzymes, which was investigated using the pnitrophenyl-α-D-glucopyranoside (PNPG) substrate. Acarbose was employed as the reference drug. In vitro analysis of all the synthesized compounds was performed in order to justify the validity of in silico outcomes. Results were found in good agreement with the in silico study. Initially, percentage inhibition values were determined and then the compounds having 50% inhibition values or above were subjected to the micro dilution experiment (Table 1). As previously stated, 1,2-benzothiazine-N-arylacetamides 12a-m were witnessed as the most effective antidiabetic agents during the in silico analysis. Herein, during in vitro screening these derivatives were also proved effective α-glucosidase inhibitors compared to 1,2-benzothiazine-N-arylacetamides 11a-m. Among all the compounds, 11c, 12a, 12d, 12e, and 12g showed the most promising α-glucosidase inhibition with IC 50 values of 30.65, 18.25, 20.76, 35.14, and 24.24 µM, respectively (Table 1). These inhibitors have IC 50 values even less than the standard drug, acarbose (58.8 µM [42,43]). These compounds showed more potent α-glucosidase inhibitions compared to the previously reported sulfonamides [63], dihydroxy pyrrolidines [64], and thiobarbituric acid enamine derivatives [65]. The compounds 11h, 11i, and 12m having IC 50 values 47.13, 40.41, and 58.45 µM, respectively, were witnessed as the moderate antidiabetic agents while other derivatives failed to show any promising inhibitory effects against α-glucosidase enzymes and were categorized as weak inhibitors or in-active compounds. It is concluded from the above discussion that not only the benzothiazine ring system but also the presence of the acetamide group was found indispensable for the in silico inhibition of α-glucosidase enzymes. Furthermore, it was noticed that the presence of polar as well as nonpolar groups in the structures of scaffolds played a crucial role toward the inhibition of enzymes by interacting with the selected residue. It is clear from the 2D diagrams of molecules that the interactions with the selected residues groups, such as CO, CONH, NH, SO2, Cl, Br, OH, phenyl, and aryl, all participated equally. Compounds with low rmsd values having good binding energies and interactions with the selected residues were of our interest.

α-Glucosidase Inhibition (In Vitro Analysis)
In order to assess the in silico results, all the compounds were also subjected to the in vitro inhibition of α-glucosidase enzymes, which was investigated using the pnitrophenyl-α-D-glucopyranoside (PNPG) substrate. Acarbose was employed as the reference drug. In vitro analysis of all the synthesized compounds was performed in order to justify the validity of in silico outcomes. Results were found in good agreement with the in silico study. Initially, percentage inhibition values were determined and then the compounds having 50% inhibition values or above were subjected to the micro dilution experiment (Table 1). As previously stated, 1,2-benzothiazine-N-arylacetamides 12a-m were witnessed as the most effective antidiabetic agents during the in silico analysis. Herein, during in vitro screening these derivatives were also proved effective αglucosidase inhibitors compared to 1,2-benzothiazine-N-arylacetamides 11a-m. Among Derivatives 11a-m with 3-benzoyl functionality exhibited low antidiabetic activity compared to derivatives 12a-m, which was also observed during in silico studies. In this series, the presence of deactivating groups (like Cl and NO 2 ) in the structure resulted in better enzyme inhibitions. However, herein, the presence of EWGs at the para-position of the N-phenyl ring was proved more effective compared to other sites as indicated by the IC 50 values of compounds 11c (30.65 µM), 11h (47.13 µM), and 11i (40.41 µM). Similarly, for the derivatives 12a-m, the presence of electron withdrawing groups (weak or strong) and weak electron donating groups (like CH 3 ) at the ortho-position of the N-phenyl group resulted in better antidiabetic activity as indicated by low IC 50 values of derivatives 12a (18.25 µM), 12d (20.76 µM), and 12g (24.24 µM). The derivatives 12a and 12d were found to be roughly three-fold more effective as inhibitors of α-glucosidase enzymes compared to acarbose. The increase in bioactivity with the introduction of EWGs in the benzoyl and N-phenyl moieties was also witnessed during the evaluation of previously reported 1,2-benzothiazine-N-arylacetamides as SARS-CoV-2 inhibitors [51]. In the series 12a-m, the variation in the position of nitro, chloro, and bromo substituents failed to enhance the enzymatic inhibition potential of the compounds.

General
Different instruments were used for the identification of structural features of synthesized compounds. For the determination of melting points, Stuart apparatus (Open capillary SMP3) was employed. 1 H and 13 C-NMR spectra were recorded with the help of an NMR-spectrophotometer (Bruker/Hazimin) at 600 MHz using a deuterated DMSO solvent. Chemical shift values (δ) were given in ppm (parts per million) relative to tetramethylsilane (TMS) as an internal standard. Coupling constants (J) were given in Hertz. For the progress of chemical reactions, silica-coated TLC-plates (Merk, 0.2 mm 60 HF254) were used. Spots were visualized under a UV lamp (254 nm).

Ligand Preparation
Structures of all these compounds were designed with the help of ChemDraw Ultra 12.02. In order to display these files in the MOE, structures of these compounds were saved in an MDL file (".sdf"). The 2D structure of the reference acarbose in the SDF file was retrieved from NCBI Pubchem. After this, 3D protonation and energy minimizations of all these compounds along with the reference acarbose were executed under default parameters in the MOE.

Protein Preparation
The 3D structure of Saccharomyces cerevisiae α-glucosidase (PDB ID: 2QMJ) was downloaded from the Protein Data Bank (http://www.rcsb.org/pdb; accessed on 10 October 2020). Binding interactions of the ligand (acarbose) with the protein residues were checked (Asp203, Asp542, Asp327, His600, Arg526) and then the ligand, solvent molecules, and unknown atoms from the receptor protein were removed. With the help of the MOE site finder module, the active site from the 3D structure of the protein was computed. The ligand site was found surrounded by Asp203, Tyr299, Asp327, Ile328, Asp366, Trp406, Trp441, Asp443, Met444, Arg526, Trp539, Gly541, Asp542, Asp571, Phe575, Arg598, and His600 residues. Later on, the site was occupied by dummy atoms and the protein was 3D-protonated and energy minimized with default parameters set in the MOE software (Ver.2014.0901).

Molecular Docking
To access the binding modes of all the ligands with α-glucosidase enzymes, ten conformations of all the ligands along with acarbose were docked into the selected site of enzymes with the help of the MOE dock module under the parameters fixed as, placement: triangle matcher, rescoring: London dG, refinement: forcefield and retain: 10. Finally, 2D and 3D maps of ligands were obtained. Binding modes of ligands were studied using rmsd values and docking scores of their top-ranked conformations. Results of molecular docking are depicted in Table 1, Table S1, and Figure 2, Figure 3, and Figure S1.

In Vitro α-Glucosidase Inhibition
A general spectrophotometric technique was used to perform the α-glucosidase inhibition assay with slight modifications [69]. All the compounds were dissolved in a dimethyl sulfoxide (DMSO) solvent while α-glucosidase enzymes (obtained from Saccharomyces cerevisiae (Sigma-Aldrich, Munich, Germany) were dissolved in a phosphate buffer (pH 6.8, 100 mM). Twelve and a half microliters of each test compound (dissolved in DMSO), 40 µL of enzymes (0.5 U/mL), and 120 µL of phosphate buffer (100 mM) were added in 96-well microliter plates. The plates were incubated for 5 min at 37 • C. After incubation, 40 µL of 5 mM p-nitrophenyl-α-D-glucopyranoside (Sigma Aldrich, Munich, Germany) was added in each sample solution and incubated for a further 30 min. Afterward, absorbance of released 4-nitrophenol was measured at 405 nm and 37 • C using a microplate reader. DMSO and acarbose (Sigma-Aldrich, Munich, Germany) were used as a negative control and reference inhibitor, respectively. The experiment was repeated thrice and inhibition percentages were calculated using the following formula.
where 'A' stands for absorbance. Finally, IC 50 values were also determined for the compounds having 50% or above inhibition values.

Conclusions
Novel 1,2-benzothiazine-N-arylacetamides have been discovered as inhibitors of αglucosidase enzymes for the first time. The 1,2-benzothiazine-N-arylacetamides 12a-m proved better inhibitors compared to derivatives 11a-m. The compounds 11c, 11h, 11i, 12a, 12d, 12e, 12g, and 12m exhibited low IC 50 values compared to acarbose (IC 50 = 58.8 µM). However, the derivatives 11c, 12a, 12d, 12e, and 12g were envisioned as the most potent inhibitors as these templates showed good docking scores, low rmsd values and significant interactions with the targeted receptor pockets during in silico experiments. Theses derivatives also showed low IC 50 values of 30.65, 18.25, 20.76, 35.14, and 24.24 µM, respectively, during in vitro α-glucosidase inhibition. We hope the present work will benefit researchers to develop more active antidiabetic agents with no side effects.