Introduction

The Enicostemma littorale Blume (E. littorale) plays a critical role in human wellbeing. Parts of the plant E. littoral were used historically in therapeutic applications against malaria, skin disorders, leprosy, diabetes, etc. This plant's constituents were beneficial therapeutic compounds because they had low toxicity, environmental friendliness, a long shelf life, and no side effects (Murali et al. 2002; Upadhyay and Goyal, 2004; Vasu et al. 2005). It is a noble source of iron, potassium, sodium, calcium, magnesium, silica, chloride, sulphate, phosphate, and vitamins B and C (Maroo et al. 2003, 2002; Sonawane et al. 2010; Thirumalai et al. 2011).

Numerous phytoconstituents have been isolated from the plant, E. littorale. The aerial sections of the plant yielded 34% of the dry alcoholic extract and 15.7% of the ash (Patel et al. 2009; Sadique et al. 1987; Sanmugarajah 2013). It has been stated in the literature that this plant produces five alkaloids, two sterols, and volatile oils(Selvaraj et al. 2014; Vishwakarma et al. 2010). Another sapogenin, betulin, was also isolated from this plant (Indumathi et al. 2014). Monoterpene alkaloids such as enicoflavin, gentiocrucine and seven diverse flavonoids have been extracted from the alcoholic extract and the structures have been categorised as apigenin, genkwanin, isovitexin, wertisin, saponarin, 5-o glucosylwertisin and 5-o glucosylisowertisin have also been isolated by Goshal et al. (1974). For the first time in this species, the occurrence of catechins, saponins, steroids, sapogenin, triterpenoids, flavonoids, and xanthones and a new flavonous C-glucoside called verticilliside was isolated(Jahan et al. 2009). The compound swertiamarin was isolated from E. littoral by the alcoholic extract(Alam et al. 2011; Leong et al. 2016; Patel et al. 2013; Sonawane et al. 2010; Vaidya et al. 2009a; Vishwakarma et al. 2004). There have also been six phenolic acids identified: vanillic acid, syringic acid, p-hydroxybenzoic acid, protocatechuic acid, p-coumaric acid, and ferulic acid (Abirami and Gomathinayagam 2011; Rathod and Dhale, 2013; Srinivasan et al. 2005). The methanol extract contained numerous amino acids such as l-glutamic acid, tryptophan, alanine, serine, aspartic acid, l-proline, l-tyrosine, threonine, phenylalanine, l-histidine mono-hydrochloride, methionine (Nagarathnamma et al. 2010; Sawant et al. 2011). Diabetic patients are advised to consume 2g of fresh E. littorale leaves on daily basis (Upadhyay and Goyal 2004). Therefore, E. littorale has been selected to investigate the antidiabetic potential.

Diabetes mellitus is a metabolic condition that increases the body's blood glucose, also known as diabetes (Pal 2009; Zelent et al. 2005). The hormone insulin converts blood sugar into energy-saving cells. In diabetic conditions, either the body cannot produce enough insulin or the insulin it produces cannot be used efficiently (Grewal et al. 2014; Singh et al. 2016). Two significant forms of diabetes are present; type 1 diabetes mellitus (T1DM) is an autoimmune disorder in the pancreas, where insulin is produced, the immune system targets and kills cells. Type 2 diabetes mellitus (T2DM) happens as the body becomes insulin resistant and the blood accumulates sugar (Grewal et al. 2018; Fyfe and Procter 2009).

Glucokinase is an enzyme involved in synthesising glucose into glucose-6 phosphate and serves a crucial function in glucose sensing (Charaya et al. 2018; Grewal et al. 2019). Therefore, agents induce glucokinase activation to be used to treat T2DM. The several various groups of compounds that have been discovered to cause glucokinase activation, such as benzamides (Charaya et al. 2018; Grewal et al. 2019; Li et al., 2011; Park et al. 2015), acetamides (Agrawal et al. 2013; Grewal et al. 2014), carboxamides (Grewal et al. 2014), acrylamides (Sidduri et al. 2010), benzimidazoles (Ishikawa et al. 2009), quinazolines, thiazoles (Agrawal et al. 2013), pyrimidines (Pfefferkorn et al. 2011), and urea derivatives (Castelhano et al. 2005; Grewal et al. 2020; Houze et al. 2013; Kohn et al. 2016; Murray et al. 2005; Polisetti et al. 2004; Sarabu et al. 2008).

After knowing the essential value of the activators of glucokinase in the control of T2DM (Filipski et al. 2012; Grewal et al. 2017, 2019; Grimsby et al. 2003; Matschinsky 2004; Zhang et al. 2016), we investigated the effectiveness of phytoconstituents of E. littorale as glucokinase activators, as per the literature which reports the hypoglycemic activity of this plant(Babu and Prince 2004; Maroo et al. 2002; Murali et al. 2002; Patel et al. 2009, 2012; Sonawane et al. 2010; Thirumalai et al. 2011; Upadhyay and Goyal 2004; Vaidya et al. 2009b; Vasu et al. 2005; Vijayvargia et al. 2000; Vishwakarma et al. 2010). We tried to identify the potential natural lead compounds from E. littorale as glucokinase activators through their binding mode in the allosteric site of the enzyme. The structures of all the significant phytoconstituents of E. littorale are represented in Fig. 1.

Fig. 1
figure 1

The structures of all the significant phytoconstituents of E. littorale

Material and methods

Calculation of Lipinski's rule of five

In order to further optimize the molecules, all the phytoconstituents were tested for violating the Lipinski's rule of five, Veber’s rule and the pharmacokinetic (ADMET) characteristics. The properties of all the phytoconstituents were calculated from SwissADME online tool (http://www.swissadme.ch/index.php).

Molecular docking

We conducted molecular docking (MD) on Lenovo ThinkPad T440p using PyRx-Virtual Screening Tool (Dallakyan and Olson, 2015). The structures of all the phytoconstituents and native ligand (.sdf File format) were downloaded from the National Center for Biotechnology Information PubChem (https://pubchem.ncbi.nlm.nih.gov/). The energy minimization (optimization) was performed by Universal Force Field (UFF) (Rappé et al. 1992).

A crystalline human glucokinase structure was obtained as input 1V4S from the Protein Data Bank (PDB) of RCSB (https://www.rcsb.org/structure/1V4S). 1V4S also contained the native ligand 5-(1-methyl-1H-imidazol-2-ylthio)-2-amino-4-fluoro-N-(thiazol-2-yl)benzamide that was used as a reference molecule for MD. In PyRx 0.8, Autodock vina 1.1.2 was used to conduct MD analyses of both the phytoconstituents and native ligands against the crystal structure of glucokinase (Dallakyan and Olson, 2015). With the aid of Discovery Studio Visualizer 2019, the composition of the enzyme was refined, purified, and prepared for MD (San Diego: Accelrys Software Inc. 2012). The specifications of the crystal structure and input compositions of human glucokinase used (PDB ID-1V4S) are provided in Table 1 of the PDB X-ray Structure Validation Report released on 10 August 2020. There were only 5 specific molecules in this entry, and there was one chain (Chain A). The entry comprises 3690 atoms, including 0 hydrogens and 0 deuteriums, which illustrates the need to incorporate hydrogen atoms in protein preparation processes for MD.

Table 1 The information of the crystal structure and input compositions of human glucokinase used (PDB ID-1V4S)

The MD was executed by using Vina Wizard Tool in PyRx 0.8. Molecules (PDBQT Files), both ligands and target (human glucokinase), were selected for MD. For the purpose of MD simulation, the three-dimensional grid box (size_x = 43.35 A0; Size_y = 59.36 A0; Size_z = 43.92 A0) was built using Autodock tool 1.5.6 with exhaustiveness value of 8 (Dallakyan and Olson, 2015). The active amino acids in the protein were analyzed and illuminated using Visualizer in BIOVIA Discovery Studio (version-19.1.0.18287) (San Diego: Accelrys Software Inc. 2012). The full MD process, the identification of cavity and active amino acid residues, were performed as defined by S. L. Khan et al. (Chaudhari et al. 2020; Khan and Siddiui 2020; Khan et al. 2020a, b, 2021; Siddiqui et al. 2021). The enzyme cavity is depicted in Fig. 2 with the co-crystallized ligand molecule.

Fig. 2
figure 2

The cavity of the enzyme is depicted with the co-crystallize ligand molecule (PDB ID: 1V4S)

Results

Pharmacokinetic characteristics are an important component of drug development because it enable researchers to assess the biological aspects of medication candidates. In order to establish whether or not the compound optimal for oral bioavailability, Lipinski's rule of five and Veber's rules was utilized (Table 2). All the phytoconstituents were studied for their ADMET characteristics better to grasp their pharmacokinetics profiles and drug-likeness qualities (Table 3). The ligand energies (kcal/mol), binding free energy (kcal/mol), root mean square deviation/upper bound (rmsd/ub), and root mean square deviation/lower bound (rmsd/lb) of the conformers generated of all the docked phytoconstituents are tabulated in Table 4. The active amino residues, reactive atom of ligands, bond length (A0), and type of interactions of phytoconstituents with glucokinase enzyme are depicted in Table 5. The 2D- and 3D-docking poses of all the docked molecules are represented in Figs. 3, 4, 5, 6.

Table 2 The molecular formula, Lipinski rule of five and Vebers’s rule
Table 3 The pharmacokinetic and drug-likeness properties of selected phytoconstituents
Table 4 The ligand energies (kcal/mol), binding free energy (kcal/mol), rmsd/ub, and rmsd/lb of the conformers generated of all the docked phytoconstituents
Table 5 The active amino residues, reactive atom of ligands, bond length (A0), and type of interactions of phytoconstituents with glucokinase enzyme (1V4S)
Fig. 3
figure 3

The 2D- and 3D-molecular interaction poses of native ligand and apigenin with the glucokinase enzyme

Fig. 4
figure 4

The 2D- and 3D-molecular interaction poses of ferulic acid and genkwanin with the glucokinase enzyme

Fig. 5
figure 5

The 2D- and 3D-molecular interaction poses of p-coumaric acid and protocatechuic acid with the glucokinase enzyme

Fig. 6
figure 6

The 2D- and 3D-molecular interaction poses of syringic acid and vanillic acid with the glucokinase enzyme

Where: GI, gastrointestinal; BBB, blood brain barrier; P-gp, p-glycoprotein.

Discussion

We tried to identify the potential natural lead compounds from E. littorale as glucokinase activators through the binding mode in the enzyme's allosteric site and binding free energies. In accordance with Lipinski's and Veber's rules (Table 2), many of the phytoconstituents did not demonstrated the drug-likeness characteristics and violated both the rules. Amongst all the molecules, betulin has a log P value of 8.28, which violates the Lipinski rule of 5 and indicates poor lipophilicity. An essential aspect of the compound that influences its function in the human body is lipophilicity. The compound’s Log P value shows the permeability of the drugs in the body to enter the target tissue (Krzywinski and Altman 2013; Lipinski et al. 2012). Isovitexin was found to have 7 hydrogen bond donors, which violates the Lipinski rule of 5. Saponarin has a molecular weight of 594.52 Da, 15 hydrogen bond acceptors, and 10 hydrogen bond donors, with 3 violations of the Lipinski rule of 5. It is preferable to look for substances that exceed the Lipinski limit of 500 Da, since this will only boost absorption. Still, there are several reports of relatively more significant compounds that are transported effectively through the cells. The remaining phytoconstituents, including native ligand, had fortunately not violated the Lipinski rule of 5, indicating better absorption and/or lipophilicity of the molecules. Many phytoconstituents violated the Veber's rule with total polar surface area (TPSA, should be less than 140) values and the number of rotatable bonds (which should be less than 10) that do not fall within the acceptable range for oral availability. Isovitexin, Saponarin, Swertiamarin, and Verticilliside violated the Veber’s rule.

For further optimization, all the molecules have been subjected to calculations of pharmacokinetics and drug-likeness properties. All the molecules did not show BBB penetration potential which is not favorable property for the drugs to be targeted for central nervous system. Unfortunately, many molecules did exhibited optimum log Kp (skin permeation, cm/s) and bioavailability scores. Many molecules violated the Ghose, Egan, and Muegge filters (Table 3). The molecules which displayed low GI absorption and violations of Lipinski and Veber’s rules have been eliminated from further optimization. Also, native ligand displayed low GI absorption. Therefore, only Apigenin, Ferulic acid, Genkwanin, p-coumaric acid, Protocatechuic acid, Syringic acid, and Vanillic acid have been selected to perform molecular docking studies on the GK enzyme.

A total 9 conformers were generated through MD for each molecule (Table 4). The conformer with zero rmsd/ub and rsmd/lb values has been treated as the best fit model for the glucokinase enzyme activation. The binding free energy and binding mode of the native ligand in the allosteric site of the enzyme have been considered a reference for validating the other molecules (Table 5 and Figs. 3, 4, 5, 6). Native ligand has binding free energy of − 7.2 kcal/mol and has formed 4 hydrogen bonds (3 conventional and 1 carbon-hydrogen bond) with ASP78 (2.04258 A0), LYS169 (2.60065 A0), and THR228 (2.20855 A0, 3.75081 A0). The hydrogen of a free primary amino group from the native ligand has formed a hydrogen bond ASP78, and the fluorine atom has formed a hydrogen bond with LYS169. THR228 has reacted with sulfur and fluorine simultaneously with forming one conventional hydrogen bond and one carbon-hydrogen bond. Native ligand showed electrostatic interactions with ARG85 (3.56863 A0), ASP205 (3.9455 A0), and ASP409 (3.70544 A0) through Pi-orbitals of the aromatic ring system (Fig. 3).

Apigenin (4′,5-trihydroxyflavone), a flavonoid, falls under the flavone class that is the aglycone of many naturally-occurring glycosides [(Ali et al. 2017; Baumann 2008; Salehi et al. 2019; Shukla and Gupta 2010)]. It has shown − 8.7 kcal/mol of binding free energy and formed 3 hydrogen bonds (2 conventional and 1 Pi-donor hydrogen bond) with TYR215 (2.06298 A0, 2.15918 A0), and TYR214 (3.00628 A0) (Fig. 3). It has formed two hydrogen bonds with TYR215 through hydroxyl and carbonyl oxygen atoms. One free hydroxyl group in apigenin has formed one Pi-donor hydrogen bond with TYR214 through hydrogen atom. It has shown many hydrophobic interactions due to Pi-orbitals of aromatic ring systems with VAL455 (3.86479 A0), TYR214 (5.11899 A0), PRO66 (4.99697 A0), ILE211 (5.46378 A0), VAL455 (4.74718 A0), ILE211 (4.92959 A0), VAL62 (4.80355 A0), PRO66 (5.06 A0), VAL452 (5.03778 A0), and ALA456 (4.9239 A0).

Ferulic acid is an organic compound; chemically, it is 3-methoxy-4-hydroxycinnamic acid. In plant cell walls a rich phenolic phytochemical is present covalently attached to arabinoxyls as side chains (Mathew and Abraham 2006; Wu et al. 2018). It exhibited -6.8 kcal/mol of binding free energy, which is less than native ligand, and therefore, this molecule does not possess potential to activate glucokinase enzyme. It has also formed an unfavorable donor-donor bond with ARG63 (1.56 A0) through hydroxyl hydrogen atom (Fig. 4).

Genkwanin is a monomethoxyflavone, which is a derivative of apigenin. It has been biosynthesized by apigenin in plants by methylation of the hydroxyl group at 7th position (Lee et al. 2015; Nasr Bouzaiene et al. 2016). Genkwanin has shown − 7.5 kcal/mol of binding free energy with glucokinase enzyme and possesses stable ligand energy of 206.69 kcal/mol. It exhibited hydrophobic interactions (Pi-sigma and Pi-alkyl) with ILE159 (3.78006 A0, 3.77297 A0, 5.34614 A0), VAL455 (3.87537 A0), ALA456 (4.63516 A0, 4.89221 A0, 4.7483 A0), LYS459 (5.06083 A0), VAL62 (5.08219 A0), PRO66 (5.00443 A0), and VAL452 (5.4002 A0) (Fig. 4). As it has not formed any hydrogen bond, which may result in poor activation of the enzyme.

p-Coumaric acid is a hydroxyl derivative of cinnamic acid and widely distributed in many plant species(Pei et al. 2016). It has shown − 6.4 kcal/mol of binding free energy and formed 2 conventional hydrogen bonds with TYR61 (1.86828 A0), TYR215 (2.07668 A0), whereas hydrophobic interactions (Pi-alkyl) with ILE211(4.58608 A0), VAL452 (4.8732 A0), VAL455 (4.66716 A0) (Fig. 5).

Protocatechuic acid is a type of phenolic acid that is naturally present and over 500 plants have it or its derivatives (active constituents), and these substances have different therapeutic potential. It has structural similarities with gallic acid, caffeic acid, vanillic acid, and syringic acid, which are well-known antioxidants found in foods and other items (Kakkar and Bais 2014). Protocatechuic acid has shown − 5.9 kcal/mol of binding free energy and formed 3 hydrogen bonds (2 conventional and 1 carbon-hydrogen bond) with THR65 (2.3016 A0), TYR215 (2.70823 A0), and VAL452 (3.71031 A0). It has demonstrated 2 hydrophobic bonds (Pi-sigma and Pi-Pi T-shaped) with ILE211 (3.47995 A0) and TYR214 (5.12043 A0) (Fig. 5). From these results, it can be concluded that protocatechuic acid does not have much potential to activate the glucokinase enzyme.

Syringic acid is a phenolic substance that is mostly present in fruits and vegetables. This compound is made by the shikimic acid process and is found in plants. It shows a wide variety of clinical applications in preventing diabetes, coronary disorders, cancer, ischemic stroke, etc. It can shield brain tissue from free radical injury, delay the development of diabetes, and is hepatoprotective medicine (Srinivasulu et al. 2018). It has shown -5.7 kcal/mol of binding free energy and formed 6 hydrogen bonds (4 conventional and 2 carbon-hydrogen bonds) with ASP205 (2.69294 A0), ARG85 (2.66692 A0, 2.33855 A0), LYS169 (2.52862 A0), ASP409 (3.28918 A0), ASN83 (3.59997 A0). It has formed 1 electrostatic (Pi-anion) bond with ASP78 (3.71563 A0) (Fig. 6). It demonstrated less binding free energy but exhibited a good number of hydrogen bonds, which may effectively activate the glucokinase enzyme. Vanillic acid has exhibited − 5.5 kcal/mol of binding free energy and formed 2 conventional hydrogen bonds with LEU25 (2.5621 A0), and SER373 (1.95412 A0) (Fig. 6).

Conclusion

Glucokinase is an enzyme involved in synthesising glucose into glucose-6 phosphate and serves a crucial function in glucose sensing. Therefore, agents that induce glucokinase activation could be used to treat T2DM. The Enicostemma littorale Blume (E. littorale) plays a critical role in human wellbeing. Parts of the plant E. littoral, were used historically in therapeutic applications against malaria, skin disorders, leprosy, and mostly antidiabetic activity of this plant have been reported in many literatures as well as it has been recommended in diabetic patients in Ayurveda system of medicine. The present work has been carried out to investigate the glucokinase activation potential of phytoconstituents of E. littorale through MD. All the phytoconstituents have been screened through the Lipinski rule of 5, Veber’s rule, and ADMET properties. From  this initial screening, only Apigenin, Ferulic acid, Genkwanin, p-coumaric acid, Protocatechuic acid, Syringic acid, and Vanillic acid have been selected to perform molecular docking studies on the GK enzyme.

MD is a computational research-based technique for exploring possible binding interfaces through the docking of proteins and drugs. A total of 9 conformers were generated through MD for each molecule. The conformer with zero rmsd/ub and rsmd/lb values has been treated as the best fit model for activating the glucokinase enzyme. The binding free energy and binding mode of the native ligand in the allosteric site of the enzyme have been considered the reference for the other molecules' validation. The native ligand has exhibited -7.2 kcal/mol binding free energy with useful binding mode into the enzyme's allosteric site, whereas; it has formed four hydrogen bonds with THR-228, LYS-169, ASP-78, and GLY-81. Based on these findings, the interactions of phytoconstituents have been justified. Apigenin, genkwanin, and swertiamarin exhibited − 8.7, − 7.5, and − 8.3 kcal/mol binding free energy, respectively, which indicates better enzyme activation than the native ligand. Swertiamarin has formed 08, whereas syringic acid exhibited − 5.7 kcal/mol binding affinity but has formed 06 hydrogen bonds with allosteric amino acid residues, which confirms the excellent enzyme activation by these phytoconstituents. Many antidiabetic Ayurvedic formulations contain E. littorale extract, which is already known to have therapeutic effects in diabetic patients. We identified and reported the lead phytoconstituent responsible for the antidiabetic potential. We have concluded that if we can isolate and consume the exact active phytoconstituents (glucokinase activators) from this plant, we can use them effectively to treat T2DM and by considering them as a natural lead compound, we can develop and validate more glucokinase activators.