Computational Analysis for Physicochemical Properties of Compounds in Senna auriculata Leaves Methanolic Extract to have Antidiabetic Potentials and their Molecular Interaction with α-amylase

Aims: Diabetes mellitus (DM) is chronic disorder well known for increased glucose level in blood. This disease can be controlled by inhibiting the enzyme (e.g., α-amylase) involve in carbohydrate hydrolysis. Senna auriculata leaves methanolic extract (SALME) have potential antidiabetic properties and it was also found to be safe in preclinical studies. In this study the aim was to explore the molecular interactions of α-amylase and bioactive compounds in SALME and their physicochemical properties. Methodology: Computational approach such as molecular docking and physicochemical analysis prediction was applied to understand the antidiabetic potential of natural compounds present in SALME. Results: The results showed from physicochemical analysis that out of 11 only 7 compounds are having drug like properties which are orally and intestinally better bioavailable. Furthermore, molecular docking analysis explained that three compounds (C3, C4, and C7) have lower binding Original Research Article Dukhyil; JPRI, 33(54A): 30-42, 2021; Article no.JPRI.78033 31 energy, ΔG (-8, -9.1, -9.5 kcal/mol) and better binding affinity, Ki (7.31 x 10 5 , 4.68 x 10 6 , and 9.2 x 10 6 M -1 , respectively) than the acarbose ΔG (-7.8 kcal/mol) and Ki (6.18 x 10 5 M -1 ), a well-known FDA approved medication for DM. The study also explained the binding pattern that the catalytic residue such as Asp197, Glu233 and Asp300 are involved in stabilizing the natural compounds with in the catalytic active site of target enzyme. Conclusions: From the results it has been concluded that these three compounds found in SALME have better inhibitory potential for α-amylase in comparison with acarbose. Further validation of the findings is required through molecular dynamics simulation, ADME-T study, and invitro enzyme inhibition by the purified compounds.


INTRODUCTION
Diabetes is a non-communicable chronic disorder which effects nearly 422 million people globally, responsible for 1.5 million deaths annually and predicted to negatively affect around 700 million people in 2045 [1,2]. Type 2 Diabetes mellitus (T2DM) occurs due to imbalance in carbohydrate metabolism which decreases the cellular concentration of glucose and negatively effects several other metabolic processes related to nephropathy, retinopathy, heart, fracture, Covid-19, neuro-disorder [3][4][5][6][7]. The common causes for T2DM are defects in insulin secretion destruction of beta cell in pancreas, insulin deficiency, and/or nonresponsive insulin receptors which leave the high level of glucose in blood and is the primary diagnostic parameter for hyperglycemia [8]. Oxidative stress and other environmental factors (cigarette smoking) also play a major role for the imbalance of several metabolic activities and can be interlinked with diabetes, inflammation, and cardiovascular diseases [9][10][11][12]. There is an increased economic burden of diabetes management and presumed to reach up to USD 2.5 million in 2030, which indicates an urgent need of cost-effective management and control of T2DM [9].
One method to control hyperglycemia is by inhibiting the enzyme (α-amylase) responsible for the catabolism of polysaccharides into smaller molecules such as monosaccharides [13]. Several therapeutic medications for the management of T2DM are well-proven αamylase inhibitors such as acarbose, miglitol and voglibose [14]. However, these medications have several side effects like diarrhea, gastrointestinal discomfort, hepatotoxicity, and pancreatitis [15]. Therefore, efforts made to develop novel inhibitors, of natural origin to minimize the side effects and economic burden [16,17].
Plants are well-known for good source of medicinal metabolites which can cure various human disease complications such as oxidative stress, gastric ulcer, microbial infection, inflammation, hyperglycemia, hyperlipidemia, neuro-disorders, including cancer by inhibiting the key regulatory enzymes [18][19][20][21][22][23][24][25][26][27][28][29][30][31][32]. Moreover, various medicinal properties (antioxidant, antidiabetic, anti-inflammatory, anticancer) have been reported in leaves, flowers, roots, seeds, and stem Cassia auriculata also known as Senna auriculata [33]. Recently it has been confirmed from invitro approach that methanolic extracts of Senna auriculata leaves (SALME) have great anti-diabetic potential including anti-inflammatory and antioxidant properties [34]. Beside invitro medicinal properties of SALME, Prasanth Kumar et al., also evaluated their metabolites through GC-MS analysis and reported twenty-one compounds which might be responsible for its medicinal properties [34]. Some of these compounds are available in chemical database such as Pubchem and ChemSpider. Therefore, it has been hypothesized in this study to evaluate the physicochemical properties, medicinal chemistry and to find that which of these metabolites are highly effective for antidiabetic properties by inhibiting α-amylase enzyme through several computational tools. Moreover, the molecular interactions of best bioactive metabolite have also been evaluated. This study will give an idea of best bioactive metabolite of SALME for antidiabetic potential which can be further confirmed through in-vitro and in-vivo approaches.

Hardware and Software Used
The 3-D crystallographic structure of the target protein (2QV4) was downloaded from the protein data bank database (http://www.rcsb.org/pdb/) [35]. The molecular docking was performed using PyRx-Python prescription 0.8 using Autodock-Vina with the Lamarckian genetic algorithm as a scoring function [30,36]. The interactions of molecules between best scoring ligand and protein were individually visualized and analyzed through Discovery Studio visualizer 2021 (BIOVIA) software tool [37,38]. The system used for computational study was Intel(R) Core (TM) i7-8550U CPU @ 1.80GHz, having 2.0 GHz processor including 16 GB RAM. The graphic card used in this workstation was Intel® UHD Graphics 620.

Preparation of Ligands
The ligands (natural compounds) predicted through GC-MS analysis of SALME were used in this study [34]. The ".sdf" and/or ".mol" file for 3-D structure of ligands were downloaded from PubChem database (https://pubchem.ncbi.nlm.nih.gov/) and ChemSpider database (http://www.chemspider.com/), respectively. These ligands were then energy minimized using universal force field (UFF) and density function theory (DFT and further converted to Autodock suitable ".pdbqt" file format through inbuilt OpenBabel tool in PyRx software.

Preparation of Target Protein
The co-crystallized 3-D structure of human pancreatic α-amylase enzyme (PDB Id: 2QV4) at 1.97 Å resolution with its native ligand (acarbose) was extracted from Protein data bank (PDB) database (http://www.rcsb.org/pdb/) [40,41]. The target protein was prepared for molecular docking study by removing the all heteroatom such as native ligands, and non-essential water molecules, adding hydrogens (polar olnly), calculating Gasteiger charges, and converting ".pdb" file format to ".pdbqt" format. The energy minimization and ensuring that no residues carry the non-integral charges of protein structure was performed using a built-in tool in PyRx.

Molecular Docking Study
Virtual screening was performed through the PyRx-Python 0.8 software. PyRx uses the AutoDock 4.2 and AutoDock Vina docking engine with Lamarkian Genetic algorithm method (Dallakyan and Olson 2015; Trott and Olson 2010). All the natural compounds (ligands) were individually docked with the α-amylase protein (2QV4). The grid box dimensions for target protein were selected through discovery studio visualizer and was set to 25x25x25 Å, the coordinates of grid box were centered at 12.33x48.02x25.63 Å, which was similar as discussed in previous report [42]. The docking was performed with the "exhaustiveness" set to 8. All other docking parameters were set to the default values of the software. The binding affinity (Kd) of ligands for the target protein was calculated from the Gibb's free binding energy (ΔG) using the following relation [30]: (1) In this equation universal gas constant is denoted as "R" whereas temperature is defined as "T".
The ligands with the minimum Gibb's free binding energy were selected for further analysis. The best pose of each "protein-ligand complex" was generated and analyzed using Discovery Studio 2021 (BIOVIA).

Physicochemical Properties of SALME Compounds
The reason behind most of the drugs failed during clinical trials and drug development process are now well understood that the druglike compounds have certain criteria to be followed and it is well documented in several reports [43,44]. Ninety percent of orally active medications that have accomplished the clinical phase 2 trial represents the four physicochemical parameters in specific range such as molecular weight (MW) ≤ 500, log P ≤ 5, hydrogen bond donors (HBD) ≤ 5, and hydrogen bond acceptor (HBA) less than 10. Compounds having more than 10 rotatable bonds usually have poor oral bioavailability [45]. Some of the SALME compounds reported in this current study follow the criteria of these physicochemical properties (Table 1). Our results depicted that out of 11 compounds (C1-C11) 5 compounds (C7, C8, C9, C10, and C11) have more than 500 of molecular weights. Three (C5, C6, and C9) out of eleven compounds (C1-C11) selected for this study have more than 9 rotatable bonds, one compound (C11) and standard drug acarbose (ACA) represents ≥10 HBA, acarbose also showed more than 5 HBD. Molar refractivity (MR) range considered to be from 40 to 130 for better intestinal and oral absorption [46]. The results showed that four SALME compounds (C7, C9, C10, and C11) have higher than 130 value for MR. Those compounds follow the three properties out of five properties of Lipinski rule can be acceptable for the drug-likeness [44]. Our results showed that C9, and C11 are not suitable as drug candidate.

Molecular Docking study
Computational study of molecular docking is widely used to identify the inhibitory property of various small molecules which can decrease the efforts and time of wet lab work [47,48]. In this study six compounds (C1, C3, C4, C5, C6, C7) were selected for molecular docking study to analyze the binding score (delta G) and binding affinity (Kd) for the target protein α-amylase (PDB I'd: 2QV4). We have excluded the compounds out of natural compounds identified in GC-MS analysis of SALME which are having higher molecular mass than 510 kD for molecular docking study. Before the docking of natural compounds with protein the native ligand (acarbose) was redocked and found that the redocked acarbose has almost bound to the similar residues where native ligand interacted, which validate the accuracy of results. The natural compounds were individually docked on the active site residues by removing the acarbose which generally binds as a competitive inhibitor [14,49]. The 2-D structure of α-amylase (2QV4) depicted through discovery studio visualizer and previous reports reveals that acarbose binds in the active site of protein with Trp59, Tyr62, Gln63, His101, Asn105, Ala106, Thr163, Arg195, Glu233, His299 and Asp300 by hydrogen bond. Moreover, several residues of protein such as Asp197, Ile235, His299, Leu165, Gln63, Ala198, were interacted with acarbose via van der Waals forces [50,51]. Similar amino acids such as Asp197, Glu233 and Asp300 were previously reported as key catalytic residues [42,52,53]. These catalytic residues having carboxyl group in it play a major role in catalyzing the α 1,4-glycosidic bond of polysaccharides and help in carbohydrate digestion [54]. Brayer et al. [55] analyzed that there was 10 6 -fold decline in catalytic activity of human pancreatic α-amylase enzyme by substituting Asp197 and 10 3 -fold decrease in activity by substituting Glu233 and Asp300. This showed the importance of these residues for catalytic activity of enzyme.

Binding Score Analysis
The results showed in Table 2, that binding energy of redocked acarbose to target protein is -7.7 Kcal/mol which is similar to the recent reports, where it was depicted as -7.  [56][57][58][59]. The results confirms that redocked acarbose mimic the binding pattern of reference inhibitor that is acarbose co-crystallized with target protein ( Table 3, Fig. 1A & 1C).   The redocked acarbose binds in the active site of protein with Asn105, Ala106, Thr163, Glu233 and Asp300 by conventional hydrogen bond and with Gln63 by carbon hydrogen bond. Moreover, several residues of protein such as Trp59, Leu165, Arg195, Asp197, Ala198, Ile235, and His299 were interacted with acarbose via van der Waals forces (Fig. 2A). These residues are similar to those which present for the interactions of the co-crystallized acarbose in native structure of protein. The acarbose demonstrate the similar interactions with protein as reported earlier [59]. All the compounds reported in SALME depicted the binding energy from -5.5 to -9.5 Kcal/mol ( Table 2). Three compounds in SALME coded as , respectively than the acarbose binding score (-7.7 Kcal/mol) and binding affinity (6.18 x 10 5 ). From the results it has been noted that the compounds who have more rings in their structure have better binding affinity and they are in relation with acarbose which also have 4 rings in its structure. The results of this study are in accordance with previous reports where natural compounds exhibited better activity than the standard drug against α-amylase [21,58].

Molecular Interactions Analysis
All the compounds of SALME interacted with the same catalytic active site pocket of α-amylase where acarbose get binds ( Fig. 1A & 1B). The interaction pattern of best three natural compounds (C3, C4, and C7) with target protein are represented in Fig. 2. The binding interactions of these three compounds were further compared with the binding pattern of control inhibitor (acarbose) and briefed in Table  3. It was observed that C3 and α-amylase complex was stabilized by three halogen and Ligand C3 -LYS200 (Pi-Alkyl) to stabilize the complex (Fig. 2D). The complex of amylase and C4 was stabilized by six halogen (fluorine) interactions between ASP197:OD1 -Ligand C4:F, ASP197:OD2 -Ligand C4:F, ASP197:OD3 -Ligand C4:F, HIS299:NE2 -Ligand C4:F, ASP300:OD1 -Ligand C4:F, and ASP300:OD2 -Ligand C4:F. Moreover, this complex was also stabilized by five hydrophobic Pi-Pi Stacked interactions between Ligand C4 and amino acid residues such as TRP59 and TYR62 of target protein (Fig. 2C). The Ligand C7 and target protein (amylase) was stabilized by two conventional hydrogen bond between Ligand C7:HN -THR163:OG1 and Ligand C7:N -THR163:OG1, whereas nine hydrophobic interactions was observed in stabilizing the complex (amylase-Ligand C7) at amino acid residues LEU162, THR163, LEU165, TRP59, ALA106, VAL107, and ALA198. It was noticed that hydrophobic interactions were prominent in C7 and amylase complex (Fig. 2B).
Interestingly, the amino acid residues of αamylase commonly engaged in the interaction with natural compounds (C3, C4, and C7) as well as acarbose in the catalytic active site gorge including ASP197, GLU233, whereas ASP300 was observed in C3, C4 and acarbose when interacted with protein which show that natural compounds interacted with all the specific residues for acarbose [42,52,53]. These docking results confirms that natural compounds are competitive inhibitor of α-amylase.
In this study the focus was on to analyzing that which compound (metabolite) has better binding affinity to inhibit the α-amylase by using molecular docking software tools and can be a better therapeutic cure for Diabetes mellitus Type 2. However, there are some limitations in this study that the further explanation of interactions through molecular dynamics simulation analysis such as RMSD, RMSF, Rg, SASA, MolSA, PSA etc. was not analyzed and there was no in-vitro protocol used to understand the enzyme inhibition by these metabolites.

CONCLUSIONS
From the current study it has been concluded that some natural metabolites present in SALME such as C3, C4, and C7 exhibited better inhibitory potential of α-amylase than the standard approved drug acarbose. These compounds could help in management for noninsulin dependent diabetes. Further validation of this docking study required through molecular dynamics simulations and in-vitro enzyme inhibition by the purified compounds present in SALME.

DISCLAIMER
The products used for this research are commonly and predominantly use products in our area of research and country. There is absolutely no conflict of interest between the authors and producers of the products because we do not intend to use these products as an avenue for any litigation but for the advancement of knowledge. Also, the research was not funded by the producing company rather it was funded by personal efforts of the authors.

CONSENT
It is not applicable.

ETHICAL APPROVAL
It is not applicable.

ACKNOWLEDGEMENT
The author would like to thank Deanship of Scientific Research at Majmaah University for supporting this work under project number No. R-2021-281. The author also extends his appreciation to Dr. Danish Iqbal for his valuable suggestions and support.