Chemoinformatic-aided Antidiabetic Analysis of the Therapeutic Potential of Phytoconstituents in Eremomastax speciosa Extracts

This research attempts to establish the antihyperglycemic potential of Eremomastax speciosa , a medicinal plant utilized in traditional West African diabetes therapy, through virtual simulation. While numerous reports have validated its biological potency, studies on the drug-likeness and antidiabetic properties of its compounds are limited. The in silico pharmacological, and toxicological profile of aqueous, methanolic/methylene phytochemicals from previously reported work was analyzed using Swiss ADME and Protox II online server. The docking process was performed using PyRx-0.8, coupled with AutoDock Vina. Phytochemicals that aligned with Lipinski’s rules for drugs were then subjected to a virtual docking simulation. This simulation replicated the inhibitory effects of E. speciosa phytochemicals on sodium-glucose co-transporters ( SGLT2) and α -amylase, similar to metformin, an FDA-approved antidiabetic medicine utilized as a control. Phytochemicals such as


Materials
The computational analyses were performed on a Lenovo T460 personal computer equipped with an Intel® Core™ i5-6300U CPU (2.4-2.5 GHz, 6 th generation), 16 GB RAM, and a 64-bit operating system with an x64-based processor running Microsoft Windows 10.The following software programs were utilized, including ChemDraw Ultra v.12.0,PyRx-0.8software coupled with AutoDock Vina 1.1.2,and BIOVIA Discovery Studio Visualizer v. 16

Data mining of E. speciosa phytochemicals
Information regarding the bioactive phytochemicals present in E. speciosa was obtained from a literature review conducted by Siwe et al. 16 .This review identified ten phytochemicals reported in the aqueous extract and 14 phytochemicals reported in the methanolic/methylene extract.This dataset of twenty-four putatively bioactive compounds formed the basis for the current study.

Prediction of ADME and toxicity
The SwissADME was employed to predict the physicochemical properties, ADME parameters, and drug-likeness of the identified small phytochemicals 17 .This online platform facilitates the exploration of these properties, aiding in the drug discovery process.Canonical SMILES strings generated using ChemDraw were uploaded into SwissADME to obtain predictions for physicochemical properties (e.g., lipophilicity, water solubility), pharmacokinetics, drug-likeness, and medicinal chemistry friendliness.To assess the potential in silico toxicity profile of the identified compounds, the canonical SMILES formula obtained from ChemDraw was used for analysis with Protox II 18 .

Retrieval, preparation, and identification of protein active sites
To predict the potential antidiabetic properties of the identified druggable phytochemicals from E. speciosa, we employed in silico molecular docking simulations.Metformin and acarbose, established antidiabetic drugs, were used as references to understand their mechanisms of action.Information regarding their mechanisms was retrieved from the ChEMBL database.Metformin is known to inhibit sodium-glucose co-transporter 2 (SGLT2) 19 , while acarbose acts as an α-amylase inhibitor 20 .The crystal structures of human SGLT2 (PDB ID 8HDH) and α-amylase (PDB ID 4GQR) were obtained from the Protein Data Bank (https://www.rcsb.org/).These protein structures were then optimized for docking simulations.The binding site coordinates for each enzyme were identified by analyzing the co-crystallized inhibitors within the respective protein structures, utilizing the Discovery Studio Visualizer.Docking simulations were performed for each phytochemical against SGLT2 and α-amylase.The docking scores and the most favorable binding poses for each complex were documented 21 .

Ligand preparation and molecular docking
To gain insights into the potential mechanisms of action of the identified phytochemicals, in silico docking simulations were performed.The known mechanisms of standard drugs metformin (SGLT2 inhibitor) and acarbose (α-amylase inhibitor) were retrieved from the ChEMBL database.Crystal structures of these standard drugs, along with all phytochemicals identified through GC-MS analysis, were downloaded in SDF format from PubChem.Open Babel was employed to prepare and optimize the downloaded structures for docking simulations.The PyRx platform coupled with AutoDock Vina was then utilized for docking simulations.Briefly, the phytoconstituents isolated from the methanolic and aqueous extracts of E. speciosa were docked into the active sites of human α-amylase and SGLT2.The protein structures used for docking were retrieved from the Protein Data Bank, focusing on co-crystallized forms to incorporate the bound ligand information.The specific amino acid residues constituting the active sites were selected based on the downloaded PDB structures 22 .

Data analysis
Following the docking simulations, the ligand-protein complex with the most favorable binding affinity (ΔG; kcal/mol) and pose was selected.This complex was then saved in the PDB file format for further analysis.The Discovery Studio Visualizer was employed to visualize and analyze the intermolecular interactions formed between the ligands and the target protein 23 .

RESULTS AND DISCUSSION
Tables I to VI detail the physicochemical properties of the aqueous extract (AES) and methanolic extract (MES) of E. speciosa.While some compounds were identified in both extracts, independent analysis was conducted to establish a reference library of compounds present in each.Molecular weight (MW) is a crucial physicochemical property affecting processes like absorption and interaction with targets 24,25 .Lipinski's rule of five suggests favorable drug-like properties for compounds with MW below 500 Da, logP less than 5, and a limited number of hydrogen bond donors and acceptors 13,26, 27 .All identified E. speciosa phytochemicals complied with these rules, indicating good oral bioavailability potential.The majority of identified compounds in the AES, with the exception of trilinolein (MW 879.83 g/mol, rotatable bonds 50, relative formula mass (MR) >250, and heavy atom 63), exhibited a molecular weight below 500 g/mol and an atomic weight <40.Additionally, all AES compounds possessed a topological polar surface area (TPSA) below 100 and contained a limited number of hydrogen bond acceptors and donors (less than 10 and 5, respectively).Notably, the physicochemical properties of the MES compounds followed a similar trend, with the exception of 4,22-cholestadien-3-one, which displayed a molecular weight exceeding 500 g/mol (Tables I and II).Tables III and IV present the MLOGP values for the identified phytochemicals in AES and MES.These values indicate the relative preference of a compound for partitioning between lipophilic (fatty) and hydrophilic (watery) environments.Generally, the phytochemicals in AES exhibited significant lipophilicity, with most MLOGP values falling below 4.5 (Table III).This suggests that these compounds preferentially partition into the lipophilic phase, potentially favoring absorption across cell membranes.Exceptions include trilinolein, olean-12-en-3-one, and α-amyrin.Conversely, MES displayed a trend towards higher lipophilicity, with a few compounds, such as ethyl iso-allocholate and lupenone, having MLOGP values exceeding 4.5 (Table IV).Water solubility is another critical factor for absorption and drug formulation 28 .While specific solubility values were not determined, the presence of these compounds in hydrophilic extracts suggests favorable interaction with the aqueous environment, potentially enhancing biological activity.SwissADME predictions indicated solubility ranging from low to moderate for various compounds, suggesting potential for bioabsorption and diverse therapeutic applications depending on solubility characteristics.Tables V and VI present the hydrophilicity properties of AES and MES, respectively.As expected, all extracts exhibited good solubility in water due to their hydrophilic nature.Investigating the potential for drug interactions, we evaluated the inhibitory effects of these extracts on cytochrome P450 (CYP) isoenzymes.Interestingly, none of the compounds identified in AES showed any inhibitory activity against CYP isoenzymes (Table VII).
Gastrointestinal absorption (GIA), blood-brain barrier (BBB) permeability, and P-glycoprotein (P-gp) substrate potential were evaluated using in silico tools.Additionally, all AES components displayed favorable absorption characteristics, with high GIA and the ability to permeate BBB, except for trilinolein, olean-12-en-3-one, and α-amyrin.In contrast, the analysis of MES components revealed that most exhibited significant inhibition of the CYP2C9 isoenzyme, potentially leading to drug interactions (Table VIII).Despite this concern, MES components generally demonstrated favorable ADME profiles, with high GIA and BBB permeability.Notably, ethyl iso-allocholate, lupenone, and 4,22-cholestadien-3-one deviated from this trend, displaying non-permeability to glycoprotein substrates.Generally, the compounds exhibited high predicted GIA, indicating good oral bioavailability 29 .Exceptions included trilinolein, olean-12-en-3-one, α-amyrin, and several others, suggesting these compounds may require alternative administration routes or formulation strategies for optimal bioavailability.Most compounds were predicted to be non-BBB permeant, potentially limiting their utility for central nervous system (CNS)-related conditions 30 .However, some compounds with lower molecular weight and fewer hydrogen bonds might have the potential to cross the BBB for targeted brain therapies 31 .P-glycoprotein is an efflux pump that limits intestinal absorption of certain drugs.While some E. speciosa compounds were predicted P-gp substrates, further investigation is needed to determine the practical impact on their bioavailability.

Phytochemicals GI absorption BBB permeant Pgp substrate CYP1A2 inhibitor CYP2C19 inhibitor
Analysis based on Lipinski's Rule of Five indicated that all identified compounds within the extracts adhered to this rule, violating no more than two of its criteria (Tables IX and X).This suggests favorable drug-likeness properties, implying potential for oral bioavailability.Additionally, the extracts exhibited significant bioavailability scores, further supporting their potential for drug development.With the exception of trilinolein and 4,22-cholestadien-3-one, which displayed synthetic accessibility scores of 8.46 and 6.78, respectively, all other identified compounds possessed values <6.5.Lower synthetic accessibility scores generally indicate greater ease of synthesis.Furthermore, none of the identified compounds triggered a pain alert, suggesting a favorable safety profile (Tables XI and XII).This absence of pain alerts is a promising finding for the potential development of these compounds into safe and efficacious drugs.The combined analysis using Lipinski's rule, bioavailability scores, and ProTox-II predictions suggests that the identified E. speciosa phytochemicals possess promising drug-like properties and low predicted toxicity 18 .These findings indicate their potential as safe and viable candidates for further drug development.
An evaluation of the predicted toxicity of identified compounds in both the aqueous and methanolic/methylene chloride extracts was conducted.All compounds, except N-methyl-N-[4-[2-acetoxymethyl-1-pyrrolidyl]-2-butynyl]-acetamide, exhibited relatively high LD50 values, indicating low acute toxicity.N-methyl-N-[4-[2-acetoxymethyl-1-pyrrolidyl]-2butynyl]-acetamide displayed a predicted LD50 of 41 mg/kg, placing it within Toxicity Class 2 (moderately toxic).In silico analysis also predicted potential carcinogenicity for 7-methyl-Z-tetradecen-1-ol acetate (prediction probability: 0.50) within the aqueous extract with LD50 of 3460 mg/kg.Four compounds in the aqueous extract (ethyl iso-allocholate [0.57], trilinolein [0.57], olean-12-en-3-one [0.64], and α-amyrin [0.98]) and three within the methanolic/methylene chloride extract (ethyl isoallocholate [0.57], 4,22-cholestadien-3-one [0.99], and 3-acetoxy-7,8-epoxylanostan-11-ol [0.99]) showed predicted tendencies to induce immunotoxicity (prediction probability ≥0.50) (Tables XIII and XIV).While the majority of identified compounds pose minimal acute toxicity concerns, the predicted carcinogenicity and immunotoxicity of certain compounds warrant further investigation.In particular, 7-methyl-Z-tetradecen-1-ol acetate and the compounds identified with potential immunotoxicity should be subjected to more in-depth analysis to confirm or refute the in silico predictions.Docking simulations were performed to investigate the potential interaction between α-amylase and bioactive compounds identified in E. speciosa.The results for 4-(1,5-dihydroxy-2,6,6-trimethylcyclohex-2-enyl)but-3-en-2-one (compound name) are depicted in Figure 3.This compound exhibited thirteen interactions with α-amylase, including two hydrogen bonds with ASP197 and THR163.Additional interacting residues involved in van der Waals forces included GLU233, ALA198, HIS101, TYR62, LEU165, LEU162, TRP59, TRP58, ASP300, HIS299, and ARG195.For comparison, the docking profile of acarbose, a known α-amylase inhibitor, is presented in Figure 4. Acarbose displayed a higher number and wider variety of interactions with α-amylase, including van der Waals forces, hydrogen bonds, and unfavorable donor-donor interactions.Molecular docking simulations were performed to assess the potential antidiabetic activity of E. speciosa compounds by evaluating their binding affinity to key diabetes-related targets: SGLT2 and α-amylase 32, 33 .Lower binding energy signifies stronger ligand-target interactions.All E. speciosa compounds displayed higher binding affinities for SGLT2 compared to metformin, a common antidiabetic drug.This suggests their potential as potent SGLT2 inhibitors, potentially reducing glucose reabsorption and mitigating diabetic kidney complications 34, 35 .While the compounds exhibited lower binding affinities for α-amylase compared to acarbose, another antidiabetic drug, a significant interaction was observed.This suggests their potential to inhibit α-amylase activity and regulate postprandial glucose levels, making them promising candidates for diabetes management.Plant-based drug discovery is gaining significant interest due to the potential for lower toxicity and fewer side effects compared to synthetic medications 36,37 .However, the success of a compound as a therapeutic candidate is highly dependent on its physicochemical properties 35 .Here, we investigated the drug-likeness and antidiabetic potential of E. speciosa phytochemicals.This study identified E. speciosa phytochemicals with promising drug-like properties, low predicted toxicity, and potential for antidiabetic activity through SGLT2 and α-amylase inhibition.Further in vitro and in vivo studies are warranted to validate these findings and explore their therapeutic potential for diabetes management.

CONCLUSION
In silico assessment of E. speciosa compounds revealed promising drug-like properties and broad applicability in pharmaceutical development.The identified compounds exhibited a range of hydrophobicities, suggesting their potential for various administration routes beyond oral delivery.This is further supported by the favorable bioavailability scores, placing them within the acceptable range for druggable candidates.Additionally, the in silico inhibition profiles against diabetic targets suggest a potential role for these compounds in diabetes management.However, further investigations are warranted to validate these findings.In vitro and in vivo studies are recommended to confirm the antidiabetic activity and establish the binding stability of the compounds with the identified diabetic targets.

Table I .
Physicochemical properties of phytochemicals from AES.

Table II .
Physicochemical properties of phytochemicals from MES.

Table III .
Lipophilicity properties of phytochemicals from AES.

Table IV .
Lipophilicity properties of phytochemicals from MES.

Table V .
Water solubility properties of phytochemicals from AES.

Table VI .
Water solubility properties of phytochemicals from MES.

Table VII .
Pharmacokinetics properties of phytochemicals from AES.

Table IX .
Drug-likeness and bioavailability score of phytochemicals from AES.

Table X .
Drug-likeness and bioavailability score of phytochemicals from MES.

Table XI .
Medicinal chemistry properties of phytochemicals from AES.

Table XII .
Medicinal chemistry properties of phytochemicals from MES.

Table XIII .
Toxicological profile of phytochemicals from AES.

Table XIV .
Toxicological profile of phytochemicals from MES.

Table XV .
ΔG of top E. speciosa compounds with SGLT2 and α-amylase.