Integration of in silico, in vitro and ex vivo pharmacology to decode the anti-diabetic action of Ficus benghalensis L. bark

Abstract

Background

Traditionally, Ficus benghalensis L. is used to treat metabolic disorders and is also recorded in the Ayurvedic pharmacopeia of India. The present study aimed to evaluate the anti-diabetic property of hydroalcoholic extract/fraction(s) of F. benghalensis L. bark via in silico, in vitro, and ex vivo approach.

Methods

Enzyme inhibitory activity, glucose uptake in rat hemidiaphragm, and glucose permeability, and adsorption assays were performed using in vitro and ex vivo methods as applicable. Further, the PASS was used to identify the probable lead enzyme inhibitors. The presence of predicted enzyme inhibitors was confirmed via the LC-MS. Similarly, the docking of ligands with respective targets was performed using autodock4.0.

Results

Flavonoids rich fraction possessed the highest α-amylase, and α-glucosidase inhibitory activity followed by maximum efficacy for glucose uptake in rat hemidiaphragm. Similarly, the hydroalcoholic extract showed the highest efficacy to inhibit glucose diffusion. Likewise, 3,4-dihydroxybenzoic acid was predicted for the highest pharmacological activity for α-amylase, ursolic acid for PTP1B, and apigenin for α-glucosidase inhibition respectively. The LC-MS analysis also identified the presence of the above hit molecules in the hydroalcoholic extract.

Conclusion

The analogs of 3,4-dihydroxybenzoic acid, apigenin, and ursolic acid could be the choice of lead hits as the α-amylase, α-glucosidase, and PTP1B inhibitors respectively. Additionally, the majority of secondary metabolites from the hydroalcoholic extract of F. benghalensis may be involved in enhancing the glucose uptake to support the process of glycogenesis.

Electronic supplementary material

The online version of this article (10.1007/s40200-020-00651-9) contains supplementary material, which is available to authorized users.

Introduction

Diabetes mellitus is the altered endocrine dysfunction in which either pancreas fails to secrete the insulin due to deficiency in the β-cell mass (Type 1 Diabetes Mellitus) or altered insulin sensitivity (Type 2 Diabetes Mellitus) [1]. Presently, the major portion of prescription for the pharmacotherapy of type 2 diabetes mellitus includes multiple synthetic oral hypoglycaemic agents that are oft affiliated with diverse side effects [2] reflecting the necessity of identifying the new-fangled salutary agent to cope diabetes mellitus.

Traditional medicines are one of the imperative sources to spot novel salutary agents to cope with multiple diseases [3]. Apart from it, the traditional plant-based medicines are composed of multiple secondary metabolites that act over manifold targets and may add a beneficial effect to manage polygenic conditions [4] like diabetes or obesity. Furthermore, folk medicines are comparatively safe in human consumption compared to synthetic oral hypoglycaemic agents. Additionally, the WHO has also recommended identifying new anti-diabetic agents from traditional medicinal plants [5].

Ficus benghalensis L. is a traditional medicine that is recorded in the Ayurvedic Pharmacopeia of India in the management of multiple endocrine disorders including diabetes mellitus [6]. Further, F. benghalensis bark has been reported as the composition of multiple secondary metabolites like flavonoids, polyphenols, steroids, and triterpenes which are well recognized for their anti-diabetic efficacy [7]. The previous report also suggests the anti-diabetic efficacy of the F. benghalensis bark in experimental animal models [8].

Pancreatic α-amylase catalyzes starch into α-(l-6 and 1–4) oligoglucans and other mini oligosaccharides (maltose and sucrose) [9]; are further hydrolyzed by α-glucosidase into monosaccharides like glucose or fructose [10] which gets absorbed into the systemic circulation and contributes in post-prandial hyperglycemia. Inhibition of these two enzymes is a well-accepted approach to manage post-prandial hyperglycemia via which one of the clinically approved anti-diabetic agents i.e. acarbose was established. Further, PTP1B is well acknowledged for its downregulatory task of insulin signaling and sensitivity [11] and maintaining pancreatic β-cell mass [12]. A previous study identified a few bioactives from F. benghalensis to inhibit α-glucosidase in the gastrointestinal tract and also regulate multiple proteins associated with insulin sensitivity and glucose uptake [13]. However, the effect of hydroalcoholic extract of F. benghalensis and its fraction (rich in particular phytochemistry) has not been investigated for its potency to enhance the glucose uptake and minimize the glucose diffusion process. Hence, the present study aimed to investigate the efficacy of hydroalcoholic extract/fraction(s) of F. benghalensis to inhibit α-amylase, α-glucosidase, and PTP1B and enhance the glucose uptake, glucose adsorption, and glucose diffusion process. Further, the study also attempted to identify a few lead hits as the enzyme inhibitors via cheminformatic tools.

Materials and methods

Chemicals, software, and database

Chemicals: α-amylase, α-glucosidase, acarbose, p-NPG, starch (Sigma-Aldrich). Rest of the chemicals was of analytical grade Database: ChEBI (https://www.ebi.ac.uk/chebi/), PubMed (https://pubmed.ncbi.nlm.nih.gov/), DIGEP-Pred (http://www.way2drug.com/GE/), STRING (https://string-db.org/), PubChem (https://pubchem.ncbi.nlm.nih.gov/), PASS (http://www.pharmaexpert.ru/passonline/), KEGG (https://www.genome.jp/kegg/), RCSB (https://www.rcsb.org/); PROCHECK (https://servicesn.mbi.ucla.edu/PROCHECK/) Software: autodock4 (http://autodock.scripps.edu/), Discovery Studio 2019 (Dassault Systèmes BIOVIA 2019).

Plant collection, extraction, and fractionation

The bark of F. benghalensis was collected from Belagavi, India (16°08′37.1”N 74°38′51.4″E), authenticated at ICMR-NITM, and herbarium was deposited for the same; accession number RMRC-1405. The freshly collected barks were washed, shade dried, and turned into coarse power. The coarse powder was then extracted and fractionated to obtain the fraction(s) rich in flavonoids, steroids, alkaloids, and polyphenol as explained by Khanal et al. [14] and stored in the airtight container, protecting from light at freezing temperature till further use.

LC-MS profile of the hydroalcoholic extract

The following conditions were maintained in running the sample on LC-MS 2010A (Shimadzu Japan). The C₁₈ column was used as a stationary phase and 90:10 v/v ratio of methanol: water was used (flow rate of 200 μL min⁻¹) as the mobile phase. The sample was dissolved in the mobile phase and injected (volume 5 μl) and absorbance was recorded at 254 nm. The electrospray ionization peaks were used to identify the compounds present in the sample.

In vitro α-glucosidase inhibitory activity

α-glucosidase inhibitory activity was performed as explained by Telagari & Hullatti [10] with minor modifications. Briefly, six working solutions of test samples (100-1000 μg/ml, 0.05 ml) and acarbose (20–120 μg/ml, 0.05 ml) were preincubated (25 °C for 10 min) with α-glucosidase (1 U/ml, 0.1 ml). After pre-incubation, p-NPG (0.05 ml) was added and incubated at 37 °C for 20 min followed by the addition of sodium bicarbonate (0.1 M, 2 ml) and stirred. The whole reaction mixture was filtered and the absorbance was recorded at 405 nm. An experiment was performed in triplicates and percentage inhibition was calculated using the following formula Inhibitory activity (%) = (1 − As/Ac) X 100

“Ac” and “As” are the absorbance of control and sample respectively.

In vitro α-amylase inhibitory activity

In vitro α-amylase, inhibitory activity was performed as explained by Telagari & Hullatti [10] with minor modifications. Briefly, multiple concentrations of test samples (100–1000 μg/ml, 0.25 ml) were pre-incubated at 25 °C with α-amylase (0.5 mg/ml, 0.25 ml) for 5 min followed by the addition of 20 μl of 1% soluble starch and incubated at 37 °C for 30 min. After 30 min, DNS reagent (0.5 ml) was added and the mixture was boiled for 10 min. The mixture was then cooled and 5 ml distilled water was added in each tube; filtered and the absorbance was recorded at 540 nm. All experiments were performed in triplicates and percentage inhibition was calculated using the following formula Inhibitory activity (%) = (1 − As/Ac) X 100

“Ac” and “As” are the absorbance of control and sample respectively.

Glucose uptake by isolated rat hemidiaphragm

This experiment was performed after receiving ethical clearance from the IAEC at KLE College of Pharmacy Belagavi (resolution no. KLECOP/CPCSEA-Reg,No.221/Po/Re/S/2000/CPCSEA,Res.28–12/10/2019). Glucose uptake by rat hemidiaphragm was performed by the method described by Chattopadhyay et al. [15] with minor modifications. Briefly, albino Wistar rats were fasted overnight and sacrificed with an overdose of anesthetic ether. Diaphragms were dissected out quickly with minimal trauma and divided into two halves. The rat hemidiaphragm were placed in test tubes and incubated with the sample or without a sample (control) for 30 min in the presence of insulin at 37 °C under aeration with shaking at 140 cycles/min. Experiments were performed in triplicates. After 30 min, the effectiveness of the sample to enhance glucose uptake was calculated as % glucose uptake = (1 − As/Ac) X 100

“Ac” and “As” represent the absorbance of control and test respectively.

Glucose adsorption assay

The glucose adsorption capacity of the extract/fraction(s) was performed as explained by Ou et al. [16] with minor modifications. Briefly (15.625–500 mg/ml) of extract/fraction(s) was added to glucose solution (100 mM) and the mixture was stirred and incubated for 6 h at 37 °C. After incubation, the mixture was centrifuged and glucose content was determined in the supernatant by using a glucose peroxidase diagnostic kit. Percentage of the glucose bound to the sample was calcuated as % of glucose bound to sample = (1 − As/Ac) X 100

“Ac” and “As” represent the absorbance of control and test respectively.

Glucose permeability assay

This experiment was also performed after receiving ethical clearance from the IAEC at KLE College of Pharmacy, Belagavi (resolution no. KLECOP/CPCSEA-Reg,No.221/Po/Re/S/2000/CPCSEA,Res.28–12/10/2019). The inhibition of glucose permeation was performed as explained by Gallagher et al. [17] with minor modification using the instrument (Fig. 1) explained by Dixit et al. [18]. Briefly, glucose, Tyrode’s solution, and 1 mL extract/fraction(s) (160 mg/mL) were introduced into rat jejunum (5 cm), and the appearance of glucose in the external solution was measured at 30, 60, 90, 120, 150, and 180 min. The AUC for the glucose was calculated by the trapezoidal rule.

Fig. 1.

Instrument used in glucose permeability assay

In silico identification of enzyme inhibitors and their biological spectrum

The inhibitors of α-amylase, α-glucosidase, and PTP1B from F. benghalensis were identified by querying the SMILES of secondary metabolites (retrieved from ChEBI) in PASS [19] at Pa > Pi. However, the biological spectrum of hit enzyme inhibitors was predicted by more than 90% cut-off using PASS.

The prediction in PASS was based on an algorithm with multilevel neighborhoods of atoms descriptors of the substance [19]. The probability (Pr) for each activity (j) was calculated as

$$Prj=\left(1+\left(\mathrm{S}j-\mathrm{S}0j\right)/\left(1-\mathrm{s}j\mathrm{s}0j\right)\right)/2$$

where,

$$\mathrm{S}j=\mathrm{Sin}\left(\mathrm{u}j/\mathrm{m}\right),\mathrm{s}0j=\mathrm{Sin}\left(\mathrm{u}0j/\mathrm{m}\right)$$

$$\text{uj}=\mathrm{\Sigma }i\text{ArcSin}\left\{\text{ri}\left(2\text{pij}-1\right)\right\},$$

$$\mathrm{u}0j=\mathrm{\Sigma }i\text{ArcSin}\left\{\text{ri}\left(2\text{pj}-1\right)\right\},$$

where.

m: descriptors counts of the compound to be predicted; ri = ni/(ni + 0.5/m): regulating factor; pj = nj/n: priori probability estimate for activity j; pij = nij/ni: conditional probability estimate for activity j for descriptor i; n: compounds count in the training set; ni: compounds count with descriptor i; nj: compounds number revealing activity j; nij: compounds count with descriptor i and revealing activity j.

In silico molecular docking

Ligand preparation

The predicted enzyme inhibitors were retrieved (3D .sdf) from the PubChem and converted into .pdb using Discovery studio. Ligands were minimized using mmff94 forcefield [20] and converted into .pdbqt format.

Homology modeling and target preparation

The 3D crystallographic proteins of α-amylase (PDB: 4W93) and PTP1B (PDB: 1NNY) with complete amino acid residues were retrieved from the RCSB protein data bank. However, a crystallographic protein with complete amino acid residues for α-glucosidase was not available. Hence, SWISS-MODEL [21] was used for α-glucosidase homology modeling (query sequence accession number ABI53718.1 and PDB: 5KZW). The pre-complex water and hetero-atoms from retrieved macromolecules were removed using the Discovery studio to avoid docking interference. The distribution of amino acids of each protein was visualized in the Ramachandran Plot using PROCHECK to foresee amino acid residues allocation in most favoured regions, additional allowed regions, generously allowed regions, and disallowed regions.

Ligand-protein docking

Ligands were docked with macromolecules using Lamarckian genetic algorithm via autodock4 [22] at center (x,y,z: 1.98, −24.74, −22.79) and size (x,y,z: 82.71, 77.71, 81.32) for α-glucosidase, center (x,y,z: −8.02, 21.29, −19.01) and size (x,y,z: 56.45, 72.56, 54.37) for α-amylase and center (x,y,z: 37.30, 30.96, 33.49) and size (x,y,z: 53.47, 63.12, 48.17) for PTP1B at exhaustiveness 8. After the blind docking, ligands were re-docked at the localized sites of respective targets. Ten different poses of ligands were obtained in which the pose with the least binding energy (kcal/mol) was preferred to visualize ligand-protein interaction in the Discovery studio.

Results

LC-MS profile of the hydroalcoholic extract

The LC-MS analysis was performed to confirm the presence of secondary metabolites as enzyme inhibitors which were identified via in silico approach. LC-MS analysis identified 3′, 4′, 5, 7-tetrahydroxy-3-methoxyflavone(1), 3,4-dihydroxybenzoic acid(2), 3-O-trans-p-coumaroyltormentic acid(3), 4-methoxybenzoic acid(4), alpinumisoflavone(5), amyrin(6), apigenin(7), cyclomorusin A(8), daucosterol(9), isoderrone(10), isowighteone (11), kaempferol(12), lupeol acetate(13), mucusisoflavone A(14), mucusisoflavone B(15), mucusisoflavone C(16), psoralen(17), ursolic acid(18) and wighteone(19) in the extract of F. benghalensis bark (Fig. S1). The 2D structure of bioactives with their molecular formula and weight is represented in Fig. 2.

Fig. 2.

Bioactives identified from F. benghalensis bark by LC-MS. MF: Molecular formula, MW: Molecular Weight

In vitro enzyme inhibitory activity

Fraction rich in flavonoids from F. benghalensis showed the highest α-amylase (IC₅₀: 388.39 ± 1.94 μg/ml) and α-glucosidase (IC₅₀: 316.97 ± 5.03 μg/ml) inhibitory activity. The inhibitory constant of a fraction rich in polyphenols, alkaloids, steroids, flavonoids, and the hydroalcoholic extract is summarized in Table 1. The percentage inhibition of individual concentration is represented in Fig. 3.

Table 1.

α-glucosidase and α-amylase inhibitory activity of hydroalcoholic extract/fraction(s) of Ficus benghalensis

Test samples	IC50 (µg/ml)
	α-glucosidase	α-amylase
Hydroalcoholic extract	552.33 ± 2.96	745.44 ± 8.08
Fraction rich in flavonoids	316.97 ± 5.03	388.39 ± 1.94
Fraction rich in polyphenols	443.24 ± 5.99	657.23 ± 8.29
Fraction rich in alkaloids	511.0 ± 5.85	621.97 ± 9.55
Fraction rich in steroids	519.76 ± 8.97	855.68 ± 4.70
Acarbose	78.43 ± 0.86	71.28 ± 0.45

IC₅₀: Inhibitory Concentration 50

Fig. 3.

a α-amylase, and b α-glucosidase inhibitory activity of hydroalcoholic extract/fractions of F. benghalensis

Glucose uptake assay in rat hemidiaphragm

Fraction rich in flavonoids from F. benghalensis showed the highest efficacy to enhance glucose uptake i.e. EC₅₀ 335.71 ± 10.15 μg/ml in the presence of insulin. Table 2 summarizes the effective concentration of fractions rich in flavonoids, alkaloids, polyphenols, steroids, and hydroalcoholic extract.

Table 2.

Effective concentration (EC₅₀) of hydroalcoholic extract/fraction(s) of Ficus benghalensis for glucose uptake in rat hemi-diaphragm

Test Agents	EC50 (µg/ml)
Fraction rich in flavonoids	335.71 ± 10.15
Fraction rich in polyphenols	499.20 ± 2.75
Hydroalcoholic extract	692.87 ± 3.33
Fraction rich in alkaloids	682.16 ± 3.71
Fraction rich in steroids	641.76 ± 2.63
Metformin	169.38 ± 2.44

EC₅₀: Effective concentration 50

Glucose adsorption assay

Hydroalcoholic extract of F. benghalensis showed the highest glucose adsorptive efficacy i.e. AC₅₀ 84.44 ± 1.65 μg/ml compared to fractions rich in steroids, saponins, flavonoids, polyphenols, and alkaloids. Table 3 summarizes the glucose adsorption capacity of each test agent.

Table 3.

Glucose adsorptivity of hydroalcoholic extract/fraction(s) of Ficus benghalensis bark

Test agents	AC50 (µg/ml)
Hydroalcoholic extract	84.44 ± 1.65
Fraction rich in steroids	142.35 ± 4.53
Fraction rich in polyphenols	165.47 ± 13.59
Fraction rich in flavonoids	283.77 ± 3.81
Fraction rich in alkaloids	248.45 ± 12.31
Cellulose	76.25 ± 0.67

AC₅₀: Absorptive Concentration 50

Glucose permeability assay

The AUC for diffused glucose from rat jejunum in the presence of the hydroalcoholic extract of F. benghalensis was significantly lower (p < 0.05, 0.01) compared to other fractions. Fig. 4 represents the AUC of the diffused glucose in the presence or absence of test agents’ from F. benghalensis bark.

Fig. 4.

Area under the curve of total glucose diffused over 180 min. Data are expressed in mean ± SD (n = 3). Data were analyzed by one-way ANOVA followed by Tukey’s Test for Post-Hoc analysis

In silico identification of enzyme inhibitors and their biological spectrum

Seven different metabolites i.e. daucosterol, psoralen, 3,4-dihydroxybenzoic acid, kaempferol, 4-methoxybenzoic acid, apigenin, benjaminamide were predicted as the α-amylase inhibitors in which 3,4-dihydroxybenzoic acid was predicted for highest pharmacological activity which was further predicted as chlordecone reductase inhibitor (Pa: 0.963). Likewise, ursolic acid was predicted for the highest pharmacological activity as a PTP1B inhibitor with promising pharmacological activity as an insulin promoter (Pa: 0.970). Further, apigenin was predicted for the highest pharmacological activity as an α-glucosidase inhibitor and chlordecone reductase inhibitor (Pa: 0.973). The hit biological activities for these compounds are summarized in Table 4. Similarly, identified enzyme inhibitors and their pharmacological activity and inactivity are represented in Fig. 5.

Table 4.

Probable biological activities of lead hits

Fig. 5.

a identified enzyme inhibitors from PASS b Pharmacological activity and inactivity of bioactives for each enzyme inhibition. I: compares activity within the enzyme inhibitors and II: compares the activity across the enzyme inhibitors

In silico molecular docking

The homology modeling showed the superimposition of template and protein model reflecting the addition of missing amino acid residues for α-glucosidase. Residues in most favored, additional allowed, and generously allowed regions were 671 (91.3%), 63 (8.6%), and 1 (0.1%) respectively for α-glucosidase. Residues in most favoured and additional allowed regions for α-amylase were 387 (92.1%) and 33 (7.9%) respectively. Similarly, residues in most favoured, additional allowed, and generously allowed region was 227 (89.0%), 27 (10.6%), and 1 (0.4%) respectively for PTP1B (Fig. 6).

Fig. 6.

Active site and Ramchandran Plot of (a) α-amylase and (b) α-glucosidase (Red represents the template and green represents the model) and (c) PTP1B

Daucosterol was predicted to possess the highest binding affinity (−9.3 kcal/mol) with α-amylase; however, there were no hydrogen bond interactions. Similarly, kaempferol (−8.4 kcal/mol) and apigenin (−8.5 kcal/mol) showed a similar binding affinity with α-amylase with 3 and 2 hydrogen bond interactions respectively (Table 5). Both compounds interacted with amino acid GLU233 from the active site of α-amylase. Mucusisoflavone A was predicted to have the highest binding affinity with α-glucosidase via 3 hydrogen bond interactions with TYR360 and GLU866. In contrast, 3,4-dihydroxybenzoic acid was predicted to bind α-glucosidase with −6.2 kcal/mol with seven hydrogen bond interactions via GLN118, ALA93, ASP91, CYS127, TRP126, and ILE98 (Table 6). Among 17 different phytoconstituents, 3-O-trans-p-coumaroyltormentic acid was predicted to possess the highest binding affinity with PTP1B via two hydrogen bonds with ALA217 and GLY183. Further, mucusisoflavone A interacted with PTP1B protein via five hydrogen bonds with ARG221, GLY183, LYS116, GLY220, and GLN266 in which one interaction was with the amino acid from the active site i.e. ARG221 (Table 7). The interaction of bioactives with the highest binding affinity and hydrogen bond interactions with respective enzymes is represented in Fig. 7. Similarly, the interaction of each bioactive with the respective protein is presented in the supplementary material.

Table 5.

Binding energy of bioactives with α-amylase

Ligand	Binding Affinity (Kcal/mol)	Number of hydrogen bonds	Hydrogen bond residues
3, 4-dihydroxybenzoic acid	−5.5	4	ARG398, ASP402, SER289
4-methoxybenzoic acid	−5.6	2	ARG195, ASP197
Apigenin	−8.5	2	GLN63, GLU233
Daucosterol	−9.3	–	–
Kaempferol	−8.4	3	TYR62, GLN63, GLU233
Psoralen	−6.7	1	GLN63

Table 6.

Binding affinity of bioactives with α-glucosidase

Ligand	Binding Affinity (kcal/mol)	Number of hydrogen bonds	Hydrogen bond residues
3′, 4′, 5, 7-tetrahydroxy-3-methoxyflavone	−7.7	3	GLN118, GLN121
3, 4-dihydroxybenzoic acid	−6.2	7	GLN118, ALA93, ASP91, CYS127, TRP126, ILE98
4-methoxybenzoic acid	−5.7	1	GLN118
Apigenin	−7.4	3	TRP126, ILE98, GLN118
Daucosterol	−8.6	2	PRO541, GLY550
Isoderrone	−8.9	1	TYR543
Isowighteone	−8.2	2	ARG585, VAL358
Kaempferol	−7.3	3	TRP126, GLN118
Psoralen	−7.2	2	GLY123, GLN118
Wighteone	−8.1	2	VAL544, ASP91
Amyrin	−9.6	–	–
Cyclomorusin A	−8.8	3	TYR543, PRO94, GLN118
Lupeol acetate	−8.2	2	GLU866, VAL867
Mucusisoflavone A	−10.8	3	TYR360, GLU866
Mucusisoflavone B	−9.8	2	PRO94, ALA93
Mucusisoflavone C	−9.3	1	TYR360
Ursolic acid	−9.5	1	GLY123

Table 7.

Binding energy of bioactives with PTP1B

Ligand	Binding Affinity (kcal/mol)	Number of hydrogen bonds	Hydrogen bond residues
3′, 4′, 5, 7-tetrahydroxy-3-methoxyflavone	−7	5	ARG221, GLN266, ASP48
3, 4-dihydroxybenzoic acid	−5.2	2	SER50, LYS36
3-O-trans-p-coumaroyltormentic acid	−8.7	2	ALA217, GLY183
4-methoxybenzoic acid	−5.1	2	SER205, VAL211
Alpinumisoflavone	−7.7	1	SER80
Apigenin	−7.8	2	ARG221, ASP48
Isoderrone	−7.9	2	ARG221, SER216
Isowighteone	−7.6	2	SER80, SER205
Kaempferol	−7.7	2	GLU115, ASP48
Psoralen	−6.1	–	–
Wighteone	−7.7	1	ARG221
Amyrin	−8.3	1	CYS215
Cyclomorusin A	−8.1	3	TRP179, GLN266, ARG221
Lupeol acetate	−8.3	1	ASN162
Mucusisoflavone A	−8.3	5	ARG221, GLY183, LYS116, GLY220, GLN266
Mucusisoflavone C	−8.6	3	ASP181, PRO180, GLN266
Ursolic acid	−8.7	3	ARG221, GLN266, ASP48

Fig. 7.

Interaction of (a) apigenin, (b) kaempferol acid with (1) α-amylase, (c) Mucusisoflavone A and (d) 3, 4-dihydroxybenzoic acid with (2) α-glucosidase and (e) 3-O-trans-p-coumaroyltormentic acid (f) Mucusisoflavone A with (3) PTP1B

Discussion

Identification of lead hit molecules from traditional folk medicines in the management of multiple complex diseases is a well-accepted approach that can be performed via the experimental screening of medicinal plants followed by computational prediction. Hence, in the present study, we investigated the hydroalcoholic extract/fraction(s) of F. benghalensis bark as α-amylase and α-glucosidase inhibitors via in silico and in vitro techniques. Further, the hydroalcoholic extract/fraction(s) were evaluated for their efficacy to enhance glucose uptake into the muscle (glucose uptake assay), inhibit glucose absorption from the gastrointestinal tract (glucose permeability assay), and their efficacy to adsorb glucose on their surface. Additionally, the study also confirmed the presence of secondary metabolites from F. benghalensis bark which were predicted as the enzyme inhibitors using LC-MS. Two enzymes i.e. α-amylase and α-glucosidase in the intestinal border linings are involved in the cleavage of 1,4 glycosidic linkage of polysaccharides and convert them into monosaccharides contributing to postprandial hyperglycemia [10]. In the present study, we identified 18 different phytoconstituents to possess the biological spectrum as α-glucosidase and 7 compounds as α-amylase inhibitors. Among them, apigenin (MF: C₁₅H₁₀O₅; MW: 270.24 g/mol) and kaempferol (MF: C₁₅H₁₀O₆; MW: 286.24 g/mol) were predicted to possess the highest pharmacological activity to inhibit the α-glucosidase enzyme. Likewise, the PTP1B is involved in the downregulation of insulin secretion [23] inhibition of glucose uptake in skeletal muscles [24]. Similarly, in the present study, the efficacy of the fraction rich in flavonoid to enhance the glucose uptake was also observed. Further, the hydroalcoholic extract was more effective in adsorbing the glucose on their surface and limited the crossing of free glucose levels from the gastrointestinal tract. As well, daucosterol (MF: C₃₅H₆₀O₆; MW: 576.8 g/mol), mucusisoflavone A (MF: C₄₀H₃₂O₁₀; MW: 672.7 g/mol) and ursolic acid (MF: C₃₀H₄₈O₃; MW: 456.7 g/mol) and 3-O-trans-p-coumaroyltormentic acid (MF: C₃₉H₅₄O₇; MW: 634.8 g/mol) were predicted to possess a highest binding affinity with α-amylase, α-glucosidase, and PTP1B respectively.

In LC-MS analysis, although three peaks i.e. 109, 444, and 533 were reflected, bioactives of F. benghalensis from the ChEBI database with identical molecular weights were not identified. This could be due to two main reasons, either the database has not recorded the secondary metabolites from F. benghalensis with molecular weight 109, 444, and 533 or these could be the metabolites’ peak of a corresponding higher molecular weight compound; for example peak 553 could be due to the cumulative metabolites of fragmented bioactives with 533+ molecular weights and the same could be for 109 and 444. Since the data recorded by ChEBI from traditional medicines were considered as the standard source of bioactives, we define the above peaks as the secondary fragments of the higher molecular weight compounds of 109, 444, and 533.

In one of the previous studies, we identified three phytoconstituents from F. benghalensis as α-glucosidase inhibitors [13] using BindingDB [25] based on the principle “similar compounds tend to bind the same proteins”. However, in the present study, the PASS predicted 18 different compounds as α-glucosidase inhibitors. This variation in the number of identified inhibitors could be due to the different algorithms of training sets; BindingDB functions over the similarity index of the predeposited compound and PASS functions over the pharmacological activity over the training sets. Due to these variabilities, we performed in silico molecular docking to confirm the findings of both the cheminformatic database. On assessing the pharmacological activity of α- glucosidase inhibition, docking results of apigenin and kaempferol were quite similar compared to mucusisoflavone A which also scored the highest binding affinity via molecular docking. Further, we attempted to identify the α-amylase inhibitors using BindingDB based on an earlier protocol [13] which failed to point a molecule as its inhibitor. Although daucosterol scored highest binding affinity with an α-amylase enzyme, the minute pharmacological activity score was predicted which could be the outcome of zero hydrogen bond interaction. However, one of the previous studies isolated daucosterol from Swertia longifoliaṣ and reported it to be a potent molecule as an α-amylase inhibitor [26]. Similarly, Dehghan et al. isolated and characterized daucosterol from the roots of Rheum turkestanicum and reported its α-amylase inhibitory activity similar to the acarbose [27]. These results suggest the probable role of daucosterol in the inhibition of α-amylase in the present study. Although 3-O-trans-p-coumaroyltormentic acid was predicted to possess the highest binding affinity with PTP1B, the pharmacological activity for PTP1B inhibition was comparatively low compared to ursolic acid which was utilized as the PTP1B inhibitor as reported previously [23]. However, the inhibitory activity of PTP1B by 3-O-trans-p-coumaroyltormentic acid needs to be further confirmed via the in vitro technique to compare its inhibitory concentration with ursolic acid.

Previously, the efficacy of flavonoids to inhibit α-amylase, α-glucosidase, and PTP1B has been reported [28, 29]. Similarly, in the present study, we identified the majority of the bioactives from F. benghalensis to inhibit these enzymes in which apigenin and kaempferol (flavonoids category) were common. Further, another class of phytochemistry i.e. triterpenoids is also well acknowledged in the inhibition of α-amylase, α- glycosidase [30], and PTP1B [31]. Likewise, in the present study, we identified three terpenoids i.e. ursolic acid, lupeol acetate, and amyrin to contribute to inhibit α-glucosidase and PTP1B.

Most of the dietary fibers have an affinity to adsorb glucose on their surface; one of the beneficial effects to inhibit the absorption of glucose from the gastrointestinal tract [16]. Hence, to assess this hypothesis, we tested the effect of hydroalcoholic extract/fractions of F. benghalensis on glucose adsorptive and glucose diffusion; identified the hydroalcoholic extract to be more effective in this mechanism. The main reason for this observation could be due to the presence of dietary fibers present in the hydroalcoholic extract but not in the other fractions. Further, the reason for the inhibition of glucose transport from rat jejunum could be due to the inhibition of GLUT transporters in rat jejunum. However, our prediction of the biological spectrum for this transporter failed to identify any hit molecule which needs to be further investigated. One of the limitations of the present study is that our findings are only based on the in silico, in vitro and ex vivo approaches which need to be further confirmed via the in vivo animal study. Further, identification of PTP1B inhibitors was based on computer simulations that needs to be further confirmed via in vitro enzyme inhibitory assay.

Conclusion

The present study evaluated the hydroalcoholic extract/fraction(s) of F. benghalensis bark for the inhibition of α-amylase, α-glucosidase, and PTP1B which could contribute to the management of diabetes mellitus. Further, fraction rich in flavonoids was identified to possess the highest α-amylase and α-glucosidase inhibitory activity and also enhanced the glucose uptake in rat hemidiaphragm. Furthermore, the hydroalcoholic extract was identified to reduce the glucose absorption from the intestinal lining.

Abbreviations

AUC

area under the curves

DNS

Dinitrosalicylic acid

EC50

Effective concentration 50

GLUT

Glucose transporter

IAEC

Institutional Animal Ethics committee

ICMR-NITM

Indian Council of Medical Research - National Institute of Traditional Medicine

LC-MS

Liquid chromatography-mass spectrometry

MF

Molecular formula

MW

Molecular weight

Pa

Pharmacological activity

PASS

Prediction of Activity Spectra for Substances

Pi

Pharmacological inactivity

p-NPG

4-Nitrophenyl-β-D- glucopyranoside

PTP1B

Protein Tyrosine Phosphatase 1B

RCSB

Research Collaboratory for Structural Bioinformatics

SMILES

Simplified molecular-input line-entry system

WHO

World Health Organization

Funding

This work has not received any funds from national and international agencies.

Data availability

Data will be provided in case of a request.

Compliance with ethical standards

Conflict of interest

There are no conflicts of interest to declare.

Ethical statement

Glucose uptake and permeability assayes were performed after receiving ethical clearance from Institutional animal ethical clearance (IAEC) at KLE College of Pharmacy, Belagavi (resolution no. KLECOP/CPCSEA-Reg, No.221/Po/Re/S/2000/CPCSEA, Res.28–12/10/2019).

Consent for publication

Not Applicable