Molecular Docking of Isolated Alkaloids for Possible α-Glucosidase Inhibition

Diabetes mellitus, one of the most common endocrine-metabolic disorders, has caused significant morbidity and mortality worldwide. To avoid sugar digestion and postprandial hyperglycemia, it is necessary to inhibit α-glucosidase, a digestive enzyme with an important role in carbohydrate digestion. The criteria for the selection of alkaloids are based on their in vitro and in vivo activities on glucose modulation. The current study assessed the bonding potential of isolated alkaloids with the targeted protein. For this purpose, the 3D structure of the target protein (α-glucosidase) was reproduced using MODELLER 9.20. The modeled 3D structure was then validated and confirmed by using the RAMPAGE, ERRAT, and Verify3D online servers. The molecular docking of 32 alkaloids reported as α-glucosidase inhibitors, along with reference compounds (acarbose and miglitol), was done through MOE-Dock applied in MOE software to predict the binding modes of these drug-like compounds. The results revealed that nummularine-R and vindoline possess striking interactions with active site residues of the target protein, and were analogous to reference ligands. In conclusion, the current study provided a computational background to the α-glucosidase inhibitors tested. This novel information should facilitate the development of new and effective therapeutic compounds for the treatment of diabetes mellitus.


Introduction
The digestive enzyme, α-glucosidase, has an important role in carbohydrate digestion and is responsible for the biosynthesis of glycoproteins. Several α-glucosidases can not only perform the hydrolysis of oligosaccharides and artificial α-glycosides with α-glycosidic bonds, but can also hydrolyze α-glucans such as glycogen and water-soluble starch [1][2][3]. α-glucosidase is the primary enzyme for digestion of carbohydrates in the small intestine. α-glucosidase is different from β-glucosidase because it acts on the 1,4-α bond [4][5][6][7]. For cellular growth and development in plants, glucose produced by the activity of these enzymes is used as a major energy source [8]. These enzymes

Homology Modeling
The homology model of the target protein, α-glucosidase, was accomplished with MODELLER 9.20. The 3D structure of the protein was modeled for further docking studies. The 3D modeled structure of the protein is shown in Figure 3.

Homology Modeling
The homology model of the target protein, α-glucosidase, was accomplished with MODELLER 9.20. The 3D structure of the protein was modeled for further docking studies. The 3D modeled structure of the protein is shown in Figure 3.

Validation of the Modeled Structure
The modeled structure validated by RAMPAGE, showing stereo-chemical verification, and the verified 3D structure by ERRAT, are shown in Figure 4. For protein structure verification, a Ramachandran plot was drawn with MOE, as shown in Figure 5.

Validation of the Modeled Structure
The modeled structure validated by RAMPAGE, showing stereo-chemical verification, and the verified 3D structure by ERRAT, are shown in Figure 4. For protein structure verification, a Ramachandran plot was drawn with MOE, as shown in Figure 5.

Homology Modeling
The homology model of the target protein, α-glucosidase, was accomplished with MODELLER 9.20. The 3D structure of the protein was modeled for further docking studies. The 3D modeled structure of the protein is shown in Figure 3.

Validation of the Modeled Structure
The modeled structure validated by RAMPAGE, showing stereo-chemical verification, and the verified 3D structure by ERRAT, are shown in Figure 4. For protein structure verification, a Ramachandran plot was drawn with MOE, as shown in Figure 5.

Preparation of Protein and Molecular Docking
Proteins were prepared for molecular docking by 3D protonation, energy minimization and prediction of active site for ligands, keeping the parameters at their defaults. Next, ligands were docked with the target protein (α-glucosidase) while using MOE software. The docking results suggested that Nummularine-R was the most potent of the tested compounds, with a docking score of −14.5691 followed by Vindoline with a docking score of −13.2250. In addition to these two compounds, Conophyline, Epiberberine, glutamic acid, and mahanimbilylacetate also showed favorable results. The docking results, along with ligand structure and their properties, are shown in Table 1.

Preparation of Protein and Molecular Docking
Proteins were prepared for molecular docking by 3D protonation, energy minimization and prediction of active site for ligands, keeping the parameters at their defaults. Next, ligands were docked with the target protein (α-glucosidase) while using MOE software. The docking results suggested that Nummularine-R was the most potent of the tested compounds, with a docking score of −14.5691 followed by Vindoline with a docking score of −13.2250. In addition to these two compounds, Conophyline, Epiberberine, glutamic acid, and mahanimbilylacetate also showed favorable results. The docking results, along with ligand structure and their properties, are shown in Table 1.      Coptis chinensis (Rhizome) Jatrorrhizine 6 In vitro anti-diabetic −9.3385 [3,4] Jatrorrhizine 6 In vitro anti-diabetic −9.3385 [3,4] Coptis chinensis (Rhizome) Biomolecules 2019, 9, 544 8 of 17 Coptis chinensis (Rhizome)

Discussion
Molecular docking is frequently used to predict the binding orientation of small molecules and drug candidates to their protein targets in order to predict their affinity and activity [28,29]. In this study, we modeled various alkaloids isolated from different plants and known for their inhibition of α-glucosidase through molecular docking.
Glucose control is an effective and long-lasting treatment for type II diabetes mellitus, minimizing both cardio-vascular and nervous system symptoms associated with the disease [30,31]. α-glucosidase inhibitors are usually recommended for diabetic patients to decrease postprandial hyperglycemia caused by the breakdown of starch molecules in the small intestine [32]. The use of plants or plant-based substances may be a suitable source of α-glucosidase inhibitors because of their low price and comparatively greater safety, with a low frequency of serious gastrointestinal side effects [33].
The docked alkaloids also exhibit inhibitory potential against other hydrolase enzymes in the

Discussion
Molecular docking is frequently used to predict the binding orientation of small molecules and drug candidates to their protein targets in order to predict their affinity and activity [28,29]. In this study, we modeled various alkaloids isolated from different plants and known for their inhibition of α-glucosidase through molecular docking.
Glucose control is an effective and long-lasting treatment for type II diabetes mellitus, minimizing both cardio-vascular and nervous system symptoms associated with the disease [30,31]. α-glucosidase inhibitors are usually recommended for diabetic patients to decrease postprandial hyperglycemia caused by the breakdown of starch molecules in the small intestine [32]. The use of plants or plant-based substances may be a suitable source of α-glucosidase inhibitors because of their low price and comparatively greater safety, with a low frequency of serious gastrointestinal side effects [33].
The docked alkaloids also exhibit inhibitory potential against other hydrolase enzymes in the same class. Previous reports [21] shows that alkaloids such as berberine and palmatine inhibit aldose reductase activity at lower concentrations, losing their inhibitory potential at higher concentrations. On the other hand, coptisine, epiberberine, and groenlandicine showed reasonable inhibitory potential. Nigelladines A-C, pyrroloquinoline and nigellaquinomine have shown strong protein tyrosine phosphatase inhibitory activity [25].
Among the 10 differencing docking alkaloids, nummularine-R was found to be the most potent of the tested compounds with a docking score of −14.5691, followed by Vindoline with a docking score of −13.2250. Both these compounds have good inhibitory activity and their docking score is in the region of certain standard ligands, such as miglitol (−15.4423) and acarbose (−14.7983). Furthermore, both these ligands exhibited a good interaction with α-glucosidase. The most potent ligand, nummularine-R, formed four hydrogen interactions with the Gln121, Met122, Arg331, and Gly546 active amino acid residues. Gln121 was observed to make a polar hydrogen bond, with the oxygen atom double bonding with the piperidine moiety of the ligand. Arg331 formed polar hydrogen bonds with the Nitrogen atom of the pentene ring of the ligand. Met122 showed acidic hydrogen interactions, whereas Gly546 showed basic hydrogen interactions with the oxygen atom double bonding with the piperidine moiety of the same ligand. These interactions are shown in Figure 6A. Ala93, Ala97, Gln121, and Trp126 formed three Hydrogen and one arene-arene interaction with the ligand. Ala93 showed greasy hydrogen interaction with the nitrogen of the benzene ring, with a bond length of 3.07 Å, while Gln121 showed greasy hydrogen interaction with the hydrogen of the benzene ring, with a bond length of 3.01 Å. Ala97 demonstrated a basic hydrogen bond with an oxygen atom double bonding with the piperidine moiety of the ligand with a bond length of 3.07 Å. Trp126 exhibited an arene-arene bond with a bond length of 3.89 Å with the benzene ring of the inhibitor. All these interactions are shown in Figure 6B. The reference compound Miglitol revealed four hydrogen bonds with Met122, Arg275 and Arg331 of the target protein active site residue. Met122 formed two acidic hydrogen bonds with the hydrogen and oxygen of the same phenol moiety, while Arg331 bound via polar hydrogen interaction The reference compound Miglitol revealed four hydrogen bonds with Met122, Arg275 and Arg331 of the target protein active site residue. Met122 formed two acidic hydrogen bonds with the hydrogen and oxygen of the same phenol moiety, while Arg331 bound via polar hydrogen interaction with the oxygen of the benzene ring of the phenol moiety of the ligand. Arg275 formed a polar hydrogen bond with the oxygen atom, double bonding with the piperidine moiety of the same compound. All these bonds are shown in Figure 6C. Another reference compound, Acarbose, showed seven hydrogens, Ala93, Ile98, Gln121, Met122, Arg275, Pro545 and one arene-cation interaction with Trp126 of the target protein. Ala93 and Met122 formed greasy and acidic hydrogen bonds with the nitrogen of the benzene ring. Met122 also formed acidic hydrogen with the carbon backbone of the benzene ring. Gln121 formed a greasy hydrogen interaction with the carbon of the benzene backbone. Arg275 exhibited a polar hydrogen bond with the OH group of the phenol moiety, similarly Pro545 possessed greasy hydrogen and Ile98 formed basic hydrogen interactions with the target protein.
Trp126 exhibited an arene-cation bound with the carbon backbone of the ligand. All these bonds are shown in Figure 6D. The 3D interaction of the most potent ligands (nummularine-R and Vindoline) and standard (Acarbose and Miglitol) with target protein are shown in Figures 7-10, respectively. Similarly, these docking results were consistent and in full agreement with the in vitro anti-diabetic activity previously reported [18,19]. In addition to nummularine-R and Vindoline, Conophyline, Epiberberine, Glutamic acid and Mahanimbilylacetate also showed good interactions with the target protein, with docking scores of −12.6274, −12.9822, −12.6023, −12.9971, −12.7703, respectively. Nummularine-R, formed four hydrogen interactions, with the Gln121, Met122, Arg331, and Gly546 active amino acid residues.  Nummularine-R, formed four hydrogen interactions, with the Gln121, Met122, Arg331, and Gly546 active amino acid residues.

Conclusion
The molecular docking of 32 alkaloids isolated from various plants, along with the standard compounds acarbose and miglitol, were docked to α-glucosidase by using MOE-Dock applied in MOE software to predict the binding modes of these drug-like compounds. The results showed that nummularine-R and Vindoline possessed striking interactions with active site residues of the target protein, α-glucosidase, and were analogous to reference ligands. Taken together, the current study provides a computational background for several α-glucosidase inhibitors. Future studies should more carefully examine the clinical efficacy of these compounds, thus facilitating the development of novel resources for the treatment of diabetes mellitus. Nummularine-R, formed four hydrogen interactions, with the Gln121, Met122, Arg331, and Gly546 active amino acid residues.

Conclusions
The molecular docking of 32 alkaloids isolated from various plants, along with the standard compounds acarbose and miglitol, were docked to α-glucosidase by using MOE-Dock applied in MOE software to predict the binding modes of these drug-like compounds. The results showed that nummularine-R and Vindoline possessed striking interactions with active site residues of the target protein, α-glucosidase, and were analogous to reference ligands. Taken together, the current study provides a computational background for several α-glucosidase inhibitors. Future studies should more carefully examine the clinical efficacy of these compounds, thus facilitating the development of novel resources for the treatment of diabetes mellitus.