Skip to main content
NCBI home page
In Silico Pharmacology logoLink to In Silico Pharmacology
. 2021 Jan 3;9(1):5. doi: 10.1007/s40203-020-00065-8

In silico analysis of a potential antidiabetic phytochemical erythrin against therapeutic targets of diabetes

Madhushree M V Rao 1, T P N Hariprasad 2,✉
PMCID: PMC7779391  PMID: 33442532

Abstract

Diabetes mellitus is a multifactorial disorder characterized by a chronic elevation in blood glucose levels. Currently, antidiabetic drugs are available to counteract the associated pathologies. Their concomitant effects necessitate the investigation for an effective and safe drug aimed to diminish blood glucose levels with fewer side effects. Several researchers are taking new initiatives to explore plant sources as they are known to contain a wide variety of active agents. Hence, the present study was undertaken to study the role of natural products using in silico interaction studies. Erythrin a compound present in lichens was selected as a potential anti-diabetic agent. Molecular docking studies were carried out with 14 target proteins to evaluate its antidiabetic potential. Molecular docking analysis resulted in favourable binding energy of interaction ranging as low as − 119.676 to − 92.9545 kcal/mol for erythrin, Analogue showed the highest interactions with 3C45 (− 119.676 kcal/mol) followed by 2Q5S (− 118.398 kcal/mol), 1XU7 (− 117.341 kcal/mol), 3K35 (− 114.267 kcal/mol). Erythrin was found to fare better than the three clinically used antidiabetic compounds, metformin, repaglinide and sitagliptin. Further, the molecular interactions between erythrin and the diabetes related target proteins was established by analysing the interactions with associated amino acids. In silico pharmacokinetics and toxicity profile of erythrin using admetSAR software predicted erythrin as non-carcinogenic and non-mutagenic. The drug-likeliness was calculated using molsoft software respecting Lipinski’s rule of five. The compound was found to comply with Lipinksi rules violating only one filter criterion. The study suggested that erythrin could be a potential anti-diabetic agent.

Keywords: admetSAR, Diabetes, Docking, Erythrin, Insilico analysis

Introduction

Diabetes mellitus (DM) is a metabolic disorder generally characterized by a chronic elevation in blood glucose levels (Gupta et al. 2017). DM and its complications are one of the leading causes of mortality and morbidity (Vo et al. 2016) in the world, with around 25% of the population being affected (Salehi et al. 2019). Prolonged hyperglycemia gradually leads to retinopathy, cardiovascular disease, and neuropathy (Damián-Medina et al. 2020). DM can be classified as Type 1 autoimmune (T1DM) or insulin-dependent and Type 2 (T2DM) or noninsulin-dependent where about 90% of the people are affected by type 2 DM (Gupta et al. 2017). The number of cases reported in 2019 was about 463 million (Safitri et al. 2020) which is projected to shoot up to 642 million by 2040 (Gupta et al. 2017). Some protein plays a cardinal role in DM including, AMP-activated protein kinase, 11 β-hydroxysteroid dehydrogenases,insulin receptor substrate, dipeptidyl peptidase IV, Creactiveprotein, alpha-glucosidase, Mono-ADP ribosyltransferase-sirtuin-6, glutamine fructose-6-phosphate amidotransferase, peroxisome proliferator-activated receptor gamma, protein tyrosine phosphatases, tyrosine kinase insulin receptor, protein kinase B and insulin receptor (Rathore et al. 2016).

Current anti-diabetic drugs, metformin, sitagliptin, etc. have made substantial progress in the impeding DM by either modulating or inhibiting these proteins (Arumugam et al. 2013) but have several disadvantages such as drug resistance, acute kidney toxicity, increasing the risk of heart attack, etc. (Salehi et al. 2019; Hu and Jia 2019). Therefore, the need of the hour is to develop a safe, efficient natural alternative with fewer side effects (Pandit et al. 2010). A paradigm shift toward natural resources has found profound interest, plants; in particular, provide a viable source of novel therapeutic moieties which offer a promising alternative. Lichens are symbiotic organisms consisting of association between a fungal partner (mycobiont) and one or more photosynthetic partners (photobiont) usually either green algae or cyanobacterium or both (Thadhani and Karunaratne 2017). Lichens are considered resourceful due to their biologically active components as they demonstrate diverse biological activities like antimycobacterial, antiviral, anti-inflammatory, antipyretic (Halama and van Haluwin 2004; Huneck 1999; Rankovic et al. 2007) analgesic, antiproliferative and cytotoxic, antioxidant (Hidalgo et al. 1994) and anti-HIV properties (Neamati et al. 1997). Predominantly, in the past few years, lichens have gained significant attention as a promising candidate for their biopharmaceutical applications. Lichen extracts have been explored for their anti-diabetic property (Thadhani and Karunaratne 2017) where their inhibitory effects on α-amylase (Umeno et al. 2016; Karthik et al. 2011; Vinayaka et al. 2013; Shivanna et al. 2015; Hengameh et al. 2016; Salin Raj et al. 2014; Valadbeigi and Shaddel 2016; Valadbeigi 2016), α-glucosidase (Thadani et al. 2011; Verma et al. 2012), and PTP1B was assessed as they are the key enzymes which aid in carbohydrate metabolism (Seo et al. 2009a, b, 2011; Cui et al. 2012).

For example, Ayyappadasan Ganesan et al. determined the antihyperglycemic effect of Parmotrema hababianum established a positive, dose-dependent co-relation with blood glucose reduction (Ganesan et al. 2016). Methanolic extracts of 6 macro-lichens exhibited inhibition of amylase enzyme (Vinayaka et al. 2013). The further finding showed that four lichens were analysed for their glucosidase inhibitory property showed affirmative outcomes (Valadbeigi and Shaddel 2016; Verma et al. 2012). Lichen metabolites gyrophoric acid, lecanoric acid, and methyl orsellinate revealed PTP1B inhibitory activity (Seo et al. 2009a, b). In Parmotrema Cooperi, ethyl haematomate showed better anti-glycation activity than the standard (Choudhary et al. 2011). Cladonia humilis illustrated elevation of insulin and glycogen synthesis in hyperglycemic rats additionally encumbered gluconeogenesis and enhanced sugar tolerance in normal rats (Zhang et al. 2012). Taken together, lichens and their metabolites have emerged with greater prospects as an anti-diabetic agent.

One of the holistic approaches to discover new ligands with therapeutic applicability is with the assistance of computational tools and molecular docking. Molecular docking is a structure-based simulation strategy that predicts the interaction between targeted proteins and ligands, the affinity of the interaction is based on a scoring entity also, they assess the binding site for a given receptor (Damián-Medina et al. 2020). In light of the evidence, we have performed an in silico comparison between the natural compound erythrin and the widely used conventional drugs (sitagliptin, metformin, repaglinide) as an antidiabetic agent. We have also performed an in silico drug-likeness prediction and toxicity study of erythrin which further could be optimized. To the best of our knowledge, this is the first report to establish erythrin as a potential molecule to manage and treat DM.

Materials and methods

Selection of proteins targets

The three-dimensional X-ray crystallographic structures of 14 proteins known to play crucial importance in DM were retrieved from the Research Collaboratory for Structural Bioinformatics (RCSB) Protein Data Bank (http://www.pdb.org) and saved in.pdb format. Proteins with PDB ID:1FM9, 1IR3, 1XU7, 1ZON, 2HWQ 2Q5S, 2QMJ, 2ZJ3, 3C45, 3CTT, 3K35, 3L2M, 4A5S and 4Y14 were used as targeted diabetic receptor proteins for molecular docking experiments. The designation and role of these proteins are presented in Table 1.

Table 1.

Designation and function of therapeutic proteins of Diabetes mellitus used for the molecular docking

Sl. no Protein Protein name Function References
1 1FM9 Heterodimer of the human RXR ALPHA and PPAR GAMMA ligand binding domains bound with 9-cis retinoic acid and GI262570 and co-activator peptides Regulates adipogenesis, energy balance and biosynthesis of lipid Kumar et al. (2007)
2 1IR3 Phosphorylated insulin receptor tyrosine kinase in complex with peptide substrate and ATP analog Propagates insulin signal transduction and insulin action Ganugapati et al. (2012)
3 1XU7 Tetrameric 11B-HSD1 Activates glucocorticoid receptors by cortisol Vo et al. (2016)
4 1ZON CD11A I- domain without bound cation Regulates glucose uptake and inhibit hepatic gluconeogenesis Sivajothi and Dakappa (2014)
5 2HWQ Peroxisome proliferator-activated receptor agonists Regulates glucose metabolism and fatty acid storage by enhancing insulin action Bharti et al. (2013)
6 2Q5S PPARGAMMA bound to partial agonist NTZDPA Controls metabolism of glucose and lipids Guasch et al. (2013)
7 2QMJ N-terminal subunit of human maltase-glucoamylase in complex with acarbose Assists in release of glucose from polysaccharides Prabha et al. (2018)
8 2ZJ3 Isomerase domain of human glucose:fructose-6-phosphate amidotransferase Controls glucose flux and regulates hexosamine pathway Vo et al. (2016)
9 3C45 Human dipeptidyl peptidase IV/CD26 in complex with a fluoroolefin inhibitor Deactivates natural hypoglycemic incretin hormone GLP-1 which restores glucose homeostasis Jawla and Kumar (2013)
10 3CTT N-terminal human maltase-glucoamylase with casuarine Fast absorption of carbohydrates, Increased postprandial elevations in plasma glucose level Bharathi et al. (2014)
11 3K35 Crystal structure of human SIRT6 SIRT6 leads to lowering of blood sugar level Vo et al. (2016)
12 3L2M Pig pancreatic alpha-amylase with alpha-cyclodextrin Responsible for absorption of glucose in blood Bharathi et al. (2014)
13 4A5S Human DPP4 in complex with a noval heterocyclic DPP4 inhibitor Transforms incretins into their inactive metabolites. Rozano et al. (2017)
14 4Y14 Tyrosine phosphatase 1B complexed with inhibitor (PTP1B:CPT157633) It is a negative regulator of the insulin signalling pathway Vo et al. (2016)
Open in a new tab

Preparation of ligand

Canonical SMILES of ligand erythrin was retrieved from PubChem compound database (https://pubchem.ncbi.nlm.nih.gov/) with PubChem CID: 72946996 and three dimensional structure of the molecule was simulated using online server CORINA (http://www.mn-am.com/online_demos/corina_demo) and was saved in.pdb format. Comparative analysis of the ligand was conducted against three selected drugs obtained from PubChem database sitagliptin (PubChem CID: 4369359), metformin (PubChem CID:4091), repaglinide (PubChem CID:65981). The 2D structures of erythrin, sitagliptin, metformin and repaglinide were generated using ACD/LABS Chemsketch 2020.1.2 version C15E41 (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Two dimensional structures of ligands. a erythrin, b sitagliptin, c metformin, d repaglinide. Erythrin is the antidiabetic molecule under study while sitagliptin, metformin and repaginide are standard antidiabetic drugs

Molecular docking analysis

Molecular docking between the ligand and the receptors was carried out using iGEMDOCKv4.2. (Generic Evolutionary Method for molecular DOCKing). iGEMDOCK employs generic evolutionary algorithms (flexible docking method). Docking each ligand to the 14 target proteins was carried out at standard docking option of 70 generations, 200 population size and 2 solutions. Bond energies, such as, van der walls (VDW) interaction, hydrogen bond (H-Bond) and electrostatic energy were investigated. Docked poses were visualized using PyMOL and the interactions were ranked according to their docked energy. Docking scores of erythrin were compared to those of the three standard drugs.

Drug likeness prediction

MolSoft (MolSoft, 2007) software (https://molsoft.com/mprop/) was used to predict the drug likeness of a molecule. Overall score was based on molecular weight, total number of hydrogen bond acceptors, total number of hydrogen bond donors and logP values. Canonical SMILES data from PubChem server was used as an input data for Molsoft prediction.

Absorption, distribution, metabolism, excretion, and toxicity (ADMET) analysis

AdmetSAR, an online server was used to predict the toxicity upon consumption of compounds and asserts whether the drug follows the Lipinski Rule. Human intestinal absorption (hia), blood–brain barrier (bbb) penetration, caco-2 permeability, carcinogens toxicity and other factors were predicted. The SMILES of the ligands was submitted to the AdmetSAR program for their pharmacokinetics and toxicity properties.

Results

Molecular docking analysis

To understand the binding efficiency of erythrin to the selected diabetic proteins, molecular docking simulation was conducted using iGEMDOCK. Standard drugs sitagliptin, metformin and repaglinide ligands were used for comparison. Erythrin exhibited the lowest binding energy (− 119.676 kcal/mol) and highest affinity to 3c45, followed by 2q5s (− 118.398 kcal/mol), 1xu7 (− 117.341 kcal/mol), 3k35 (− 114.267 kcal/mol) and 1zon (− 110.167 kcal/mol). The free binding energy of all the four ligands on the 14 protein targets is presented in Fig. 2. The binding energy of erythrin to all the target proteins was found to be lower than that of the standard drugs, metformin and repaglinide. Likewise, the binding energy to eleven target proteins was also found to lower except 2Q5S, 2ZJ3, 3L2M when compared to that of sitagliptin. Molecular docking analysis resulted in favourable binding energy of interaction ranging as low as − 119.676 to − 92.9545 kcal/mol for erythrin, − 126.1 to − 87.3934 kcal/mol for Sitagliptin, − 83.3246 to − 64.8698 kcal/mol for metformin and − 109.662 to − 75.8214 kcal/mol for repaglinide (Table 2). The molecular interactions of erythrin against the therapeutic targets of DM are shown in Fig. 3a–l. For all docking analysis, highest negative scores indicate a better active compound. The results revealed that erythrin showed lesser negative scores as compared to all the three standard drugs.

Fig. 2.

Fig. 2

Open in a new tab

Binding energies of the ligands on interaction with 14 therapeutic targets of Diabetes mellitus

Table 2.

Total energy (kcal/mol) of ligands (erythrin, sitagliptin, metformin, repaglinide) against the therapeutic proteins of Diabetes mellitus

Sl. no Protein Erythrin Sitagliptin Metformin Ripaglinide
1 1FM9 − 109.07 − 95.32 − 64.8698 − 109.66
2 1IR3 − 109.221 − 102.54 − 69.558 − 90.859
3 1XU7 − 117.341 − 101.78 − 75.224 − 93.1828
4 1ZON − 110.167 − 91.04 − 66.7341 − 75.8214
5 2HWQ − 105.449 − 91.37 − 74.8974 − 92.9746
6 2Q5S − 118.398 − 126.1 − 74.8104 − 100.61
7 2QMJ − 106.953 − 95.59 − 69.3125 − 95.4608
8 2ZJ3 − 92.9545 − 97.25 − 75.2275 − 91.902
9 3C45 − 119.676 − 89.6606 − 67.0192 − 87.8929
10 3CTT − 104.88 − 88.0089 − 74.2831 − 95.35
11 3K35 − 114.267 − 107.47 − 77.2387 − 87.04
12 3L2M − 98.4878 − 104.74 − 83.3246 − 93.57
13 4A5S − 107.063 − 87.3934 − 70.1321 − 106.11
14 4Y14 − 103.654 − 91.4555 − 74.7711 − 88.41
Open in a new tab

From the interactions analysed, protein targets to which the ligands had higher affinity were represented in bold, which signifies that either the ligands could modulate or inhibit the respective protein target

Fig. 3.

Fig. 3

Fig. 3

Open in a new tab

Molecular interactions of erythrin with the therapeutic target proteins (a–l). a 1IR3, b 1FM9, c 2HWQ, d 2Q5S. e 1XU7, f 3K35, g 4Y14, h 4A5S. i 3C45, j 1ZON, k 3CTT, l 2QMJ

Our results suggest that twelve of fourteen proteins best interacted with erythrin. 3c45 exhibited the lowest binding energy (− 119.676 kcal/mol) and highest affinity forming bonds with SER-630, TYR-631,TYP-662, GLU-205/206, ARG-125, ASN-321; followed by 2q5s (− 118.398 kcal/mol) with residues GLY-284, ARG-288, SER-342, ARG-280, ILE-281, GLU-343; 1xu7 (− 117.341 kcal/mol) interacting through with ASN-123, THR-222, LEU-217, ASN-119, LEU-215, THR-220, ILE-218; 3k35 (− 114.267 kcal/mol) with amino acid residues PHE-62, ASN-112, GLN-111, THR-213, SER-214, ILE-217, LEU-184 and 1zon (− 110.167 kcal/mol) forming residues with HIS-201,LYS-200,MET-202,GLN-172,LYS-179,GLU-180,HIS-198,LEU-195,ALA-194. Binding energies and the associated amino acid interactions of erythrin with the diabetic targets is presented in Table 3.

Table 3.

Predicted active site residues of target diabetic proteins

Sl no Protein Energy Associated amino acids
1 1FM9 − 109.07 ARG-316, ALA-372, TYR-473, HIS-323, HIS-449, ILE-326, SER-289, SE-342, GLU-259, ARG-280
2 1IR3 − 109.221 GLU-1077, MET-1079, LYS-1030, ASP-1083, PHE-1007, SER-1006, ARG-1136, ASN-1137, ASP-1150, GLU-1047, ARG-1139, ARG-1155, ASP-1156, LYS-1127, GLY-1166, ARG-1164, PHE-1184, LEU-1170, GLU-1159, THR-1160, GLY-1166, PHE-1186, ARG-1131, ASP-1132, LYS-1168, GLU-1043, LEU-1171, MET-1153
3 1XU7 − 117.341 LYS-187, TYR-183, ALA-172, TYR-177, LEU-276, TYR-280, ASN-123, THR-222, LYS-44, SER-43, THR-70, GLY-41, THR-122, ASP-259, LEU-217, GLN-234, THR-92, MET-93, ASN-119, LEU-215, SER-260, TYR-257, GLY-47, ILE-46, THR-220, ILE-218, LEU-266, SER-272, ASN-270, LYS-274, ARG-273, RG-198, SER-195, LEU-23, GLU-25, GLU-254, GLU-255
4 1ZON − 110.167 HIS-201, LYS-200, MET-202, GLN-172, LYS-179, GLU-180, HIS-198, LEU-195, ALA-194
5 2HWQ − 105.449 ILE-326, ALA-292, ARG-288, CYS-285, GLY-284, SER-342, GLU-259, ARG-280, ILE-281, GLU-295
6 2Q5S − 118.398 GLY-284, ARG-288, SER-342, ARG-280, ILE-281, ILE-362, GLU-295, MET-329, LEU-228, GLY-344, GLU-343, LEU-340, GLU-259
7 2QMJ − 106.953 ILE-133, VAL-131, SER-126, GLN-92, ARG-96, GLN-117, LEU-280, HIS-115, ASN-209, GLY-208, ASN-207, THR-205, PHE-535, ALA-536, ALA-285, ALA-780, HIS-645, ASP-702, GLU-704, ASN-750, THR-747, ASN-352, GLU-305, GLU-436, TYP-301, PHE-437, ASP-327, ARG-298, HIS-600, ILE-523, TRP-406, MET-388, LYS-389, TRP-391, VAL-398
8 3C45 − 119.676 SER-87, ASN-74, ASN-80, GLN-308, GLU-309, GLU-332, LYS-554, ARG-581, ARG-596, ASP-678, ASN-219, ASN-85, ASN-150, ASN-75, LYS-267, GLN-227, MET-348, ASN-321, SER-630, TYR-631, TYP-662, ASN-710, HIS-740, GLY-714, ASP-739, GLU-738, ASP-737, THR-736, LYS-721, ASP-243, ARG-125, LYS-122, SER-209, GLU-232, GLU-205, GLU-206, GLN-553, CYS-551, ASN-229, THR-231
9 3CTT − 104.88 VAL-150, GLN-130, ASP-153, VAL-131, ILE-133, VAL-393, TRP-391, LYS-389, ASP-414, ASN-621, ARG-624, ASN-628, PRO-740, GLN-739, ALA-536, GLY-533, THR-742, ALA-285, SER-288, GLU-353, ASN-352, GLU-436, ASP-438, SER-521, ASP-772, GLU-774, GLU-312, ARG-313, GLU-309, ASP-607, ASP-542, TRP-406, ASP-203, VAL-783, ASP-443, ARG-526, ASP-327, TRP-539, ASP-571, HIS-600, ARG-298, PHE-535, GLN-318
10 3K35 − 114.267 LYS-294, THR-160, GLY-153, LYS-31, ARG-251, ARG-106, ARG-203, ARG-229, ARG-230, ARG-124, GLU-841, GLN-145, ASP-194, ASP-192, ARG-162, LEU-190, SER-189, HIS-253, LYS-79, TRP-69, HIS-66, ARG-63, PHE-62, ASP-81, HIS-131, ASN-112, GLN-111, GLU-20, GLN-240, LEU-239, ASN-238, VAL-256, GLN-216, THR-213, ASN-222, PRO-219, SER-214, THR-55, GLY-52, ALA-51, ILE-217, LEU-215, PRO-65, ARG-74, LEU-184
11 4A5S − 107.063 GLU-146, ASN-150, SER-87, ASN-80, GLU-67, ASP-678, ARG-596, ARG-597, ARG-684, ASP-679, TYR-662, TYR-631, ARG-658, ARG-125, TRP-627, ALA-564, TYR-48, TYR-662, GLU-232, ASN-229, GLN-227, GLU-205, GLU-206, ASP-739, HIS-740, SER-86, GLN-308, GLU-309, ASN-219, ARG-147, LYS-267, ASN-281
12 4Y14 − 103.654 GLU-129, GLN-123, GLU-130, PRO-89, CYS-92, ARG-47, ASP-48, SER-216, ILE-219, ARG-221, GLY-220, ALA-18, ASP-22, HIS-54, GLN-127, PHE-182, ALA-217, TYR-46, LYS-73, GLN-78, ARG-79, SER-80, SER-203, SER-205, PRO-206, HIS-208, GLY-209, VAL-211, LYS-128, GLY-218
Open in a new tab

ADMET results

The ADMET profiles of the compounds were established by admetSAR server for evaluation of pharmacokinetic and pharmacodynamic properties. ADMET properties of erythrin, metformin, repaglinide and sitagliptin which includes Caco-2 cell permeability, brain/blood barrier, human intestinal absorption, AMES toxicity and carcinogenicity were elucidated in the present study. Results of admetSAR were analysed and tabulated in Table 4. Erythrin showed human intestinal absorption. All the compounds showed close values to in silico simulation of human cell line used (Caco-2). Carcinogenicity and mutagenicity was not observed in any compounds.

Table 4.

Pharmacokinetic and Toxicological Parameters of the ligands (erythrin, sitagliptin, metformin, repaglinide)

Molecules HIA Caco-2 permeability BBB Carcinogenicity Ames mutagenesis
V P V P V P V P V P
Erythrin + 0.9648 − 0.7992 − 0.3935 − 0.8143 − 0.7300
Sitagliptin + 0.9733 − 0.6132 + 0.9821 − 0.8429 − 0.6800
Metformin + 0.9687 − 0.9372 + 0.9807 − 0.5600 − 0.6600
Retaglinide + 0.9862 − 0.7491 + 0.9449 − 0.7429 − 0.8200
Open in a new tab

HIA human intestinal absorption, BBB blood brain barrier, V value, P probability

Drug-likeness prediction

The compounds were reviewed for drug-relevant properties respecting “Lipinski’s rule of five”, which states that a molecule to be considered as a drug should possess certain properties; molecular weight should be less than or equal to 500 Da, number of hydrogen bond donor should be less than 5 and the number of hydrogen bond acceptor should be less than or equal to 10. The water-octanol coefficient ratio should be less than 5. A compound is said to comply the drug-like if violates no more than one filter criterion. Erythrin complies with the rules with one violation (HBD ≤ 5) whereas the standard compound repaglinide violates one rule (Log P ≤ 5). Drug likeliness scores for erythrin, metformin, repaglinide and sitagliptin were 0.64, − 0.82, 0.85 and 0.52 respectively. MolLogP, MolLogS, Mol PSA and MolVol are recorded and presented in Table 5 and Fig. 4a–d.

Table 5.

Molecular properties of the ligands (erythrin, sitagliptin, metformin, repaglinide)

Molecules Molecular weight Number of HBA Number of HBD MolLogP MolLogS
Log (mol/L)		 MolPSA
A2 		MolVol
A3 		Drug-likeness score
Erythrin 422.12 10 6 0.50 − 1.11 141.88 392.75 0.64
Sitagliptin 407.12 4 2 1.34 − 1.74 61.67 329.55 0.52
Metformin 129.10 2 5 − 1.00 − 0.54 69.77 119.98 − 0.82
Repaglinide 452.27 4 2 5.20 − 4.60 62.84 478.12 0.85
Open in a new tab

HBA hydrogen bond acceptors, HBD hydrogen bond donors

Fig. 4.

Fig. 4

Open in a new tab

Drug likeliness scores of erythrin along with three clinically used compounds. (a–d). a erythrin, b sitagliptin, c metformin, d repaglinide

Discussion

The management of Diabetes mellitus has continued to be a challenge all over the world. DM is an enfeeble and persistent illness outlined by multiple aetiologies including dyslipidemia. The escalating rate of DM serves a major dilemma as the available anti-diabetic drugs have either the detriments of exhibiting low efficacy or serious side effects (Bhutani et al. 2018). Owing to this, determination of biological targets for the development of new classes of antidiabetic agents with bioactive substance and novel mechanism are today’s need. Further, designing molecules instigating therapeutic intervention remains one of the main objectives. In-silico techniques serve as a persuasive tool for elucidation of preliminary information on drug likeliness and understanding mode of action, thereby conserving time and resources (Rathore et al. 2016). In the present study erythrin, a bioactive substance from lichen was subjected to its pharmacotherapy property. The objective was to decipher its underlying mechanisms as an anti-diabetic agent and to suggest as an alternative strategy to ameliorate DM. Fourteen proteins known to play a crucial role in the development of DM were selected. Alteration of these proteins either by inhibition or expression might possibly control the level of glucose in the blood and diminish the associated comorbidities. Docking analysis substantiated the activity of proteins to be amenable to alteration by interaction with erythrin. Molecular interactions, such as the number of hydrogen bonds, between the ligands and related proteins, were analyzed by Pymol Molecular Visualization system. Hydrogen-bonds are a fundamental feature in determining the specificity of ligand binding (Wade and Goodford 1989).

Insulin receptor (IR) is a tetrameric protein undergoes conformational changes upon activation by insulin. Any inequity in production or response of insulin adds to impairment in insulin signaling, leads to the onset of DM. Erythrin mimics insulin in effectively binding to the insulin receptor (IR) at Glu-1077, Met-1079, Lys-1030, Asp-1083, Asn-1137, Asp-1150, Ser-1006by hydrogen bonding interactions (Ganugapati et al. 2011) (Fig. 3a). Insulin receptor docked with banana flower flavonoids, suggested its activation and bind to similar residues mentioned (Ganugapati et al. 2012).

The peroxisome proliferator-activated receptors (PPARg) are a type 2 nuclear receptor. The upregulation of PPARg activity efficaciously ameliorates whole-body and maintains glucose homeostasis (Damián-Medina et al. 2020). Erythrin binding with Arg 288, Tyr 473, His 323, His 449, and Ser 342 act as PPARg agonists thus improve insulin sensitivity and glucose uptake (Fig. 3 b−d). There is enough evidence showing the PPARg agonist activity of several natural resources,tocopherol fraction of raw seeds of Cucurbita pepo L showed antidiabetic activity which was further supported by molecular docking analysis (Bharti et al. 2013).

11β-HSD is an NADPH dependent enzyme which catalyzes the interconversion of the glucocorticoids, cortisone, and cortisol in humans. High circulating levels of the active glucocorticoid cortisol can initiate insulin resistance causing hepatic gluconeogenesis ultimately leading to insulin-resistant diabetes and hypertension (Damián-Medina et al. 2020). As found in our study, Erythrin might play a role in decreasing11β-HSD activity by binding with Leu 217, Thr 220, Thr 222, and Asn 119 (Fig. 3e) thus, inhibiting the enzyme that antagonizes insulin-mediated glucose uptake (Rathore et al. 2016; Trinh and Le 2014). An in silico investigation on polyphenols found in blue corn and black bean extracts showed the potential to interact and modulate 11β-HSD therefore suggesting as a therapy for DM (Damián-Medina et al. 2020).

Mono-ADP ribosyltransferase-sirtuin-6 (SIRT6) is a stress-responsive protein deacetylase and mono-ADP ribosyl transferase enzyme encoded by the SIRT6 gene. The absence of SIRT6 showed drastic inducing of blood glucose level controlling the way that a cell uptakes glucose (Trinh and Le 2014). Erythrin was found to interact with Gln 111, Phe62, Ile217, Asn112, Trp 186 (Fig. 3f). These residues seem to have a critical role in the active site of SIRT6 thus, leading a way to treat diabetes (Trinh and Le 2014; Singh et al. 2019). An in silico screening of medicinal plants resulted in 6 compounds with anti-diabetic potential and showed strong binding affinity towards the receptor with the abovementioned residues as a prominent site of binding (Singh et al. 2019).

Protein-tyrosine phosphatase 1B (PTP1B) is a negative regulator of the insulin signaling pathway. Expression of PTP’s in the pancreas stimulates compensatory islet growth leading to a dysfunction of insulin secretion. In muscle and adipose tissue, leading to the development of T2D (Damián-Medina et al. 2020). Erythrin might influence the inhibition of protein tyrosine phosphate activity by binding with SER-216, ALA-217, ILE-219, and ARG-221 residues (Fig. 3g). Thus, it could amend insulin metabolism and fat accumulation. Herbal remedies have found a profound interest in the medication of DM. Various bioactive compounds in E. thymifolia and the receptor resulted in strong interaction (Vo et al. 2016).

Dipeptidyl peptidase IV (DPP-IV) is a membrane-anchored, serine protease enzyme accountable for inactivation of glucose-regulating incretin hormones [(GLP)-1, GIP] by rapidly metabolizing them. GLP-1 stimulates insulin biosynthesis, but this is short-lived because of DPPIV catalytic activity. Therefore, inhibiting DPP4 can help in regulating glycemia (Patel and Ghate 2013). In our study it was found that erythrin forms bonds with Glu205/206, Ser-630, Tyr-631, Tyr-662 influencing the inhibition of DPP-IV activity, responsible for the degradation of hormones which regulating glucose (Fig. 3h, i). An in silico assessment of G. bicolor with the receptor showed interaction with similar binding site residues, emerging as a natural DPP4 inhibitor (Rozano et al. 2017).

AMP-activated protein kinase (AMPK) is metabolic stress sensing protein kinase that stimulates glucose uptake in muscle, fatty acid oxidation in muscle and liver, and the inhibition of hepatic glucose production (Sivajothi and Dakappa 2014). In this study we found erythrin interacting with His-201, Lys-200, -202, Gln-172, Glu-180, His-198, Leu-195, Ala-194 possibly increasing the activity of AMPK (Fig. 3j). In a study conducted to evaluate the AMPK activation, activity of pyran ester isolated from T. cannabina, showed a significant reduction in the blood glucose levels in a dose-dependent manner (Sivajothi and Dakappa 2014).

α-Glucosidase is a membrane-bound enzyme that plays an important role in catalyzing the final step in the digestion of carbohydrates. Inhibition of alpha-glucosidase postprandial glucose levels can be regulated. Ligand–enzyme interaction analysis found that the interaction of erythrin with His-600, Asp-443, Ala-536, Ala-285 in the active site hence, might affect suppressing postprandial glucose level (Fig. 3k, l). A study on OPA as an anti-hyperglycemic agent showed strong inhibition of alpha-glucosidase in contrast to the commercial drug acarbose (Lee et al. 2014).

The present study was carried out to understand drug likeliness and ADME properties of erythrin. Log p is an important consideration as it is a measure of permeation of drugs across cell membranes. Moderate log P in the range 0–3, allude for good gastrointestinal absorption after oral dosage. Both erythrin and sitagliptin were in the acceptable range. Candidate molecule along with standard drugs showed good drug dissolution with aqueous solubility log S > − 6. Molecular weight of all the ligands was found to be less than 500 and thus anticipating their easy transportation, absorption, and diffusion. Hydrogen-bonding describes drug permeability. A poor permeation co-relates to more than 5 H-bond donors and 10 H-bond acceptors. Erythrin had a slightly higher permissible limit for HBA, as predicted by Molsoft. Pharmacokinetic and toxicological parameters such as BBB (Cbrain/Cblood), human intestinal absorption, plasma protein binding, mutagenic and carcinogenic effects were predicted using admetSAR. It is seen that human intestinal absorption value closer to 1 represents better absorption through the intestine. Erythrin showed a value of 0.9648 represents human intestinal permeability. Also, erythrin was non-carcinogenic and mutagenic. Thus, erythrin can emerge as a potentially promising agent for the treatment of DM.

Conclusion

The prevalence of DM and detriments of the current anti-diabetic drug demands an investigation of potent drugs preferably from the natural source. Plants with good ethnobotanical history have great scope to emerge as a safe and effective therapy in the alleviation of DM. In the present research for the first time, an attempt has been made to exploit erythrin for its anti-diabetic property. Proteins known to play a crucial role in DM were docked and the results showed therapeutic intervention of erythrin by arbitrating with multiple targets operating in diabetes. Further, drug likeliness and the ADMET profile of the compound were evaluated by molsoft and admetSAR respectively. The compound was found to comply with Lipinksi rules violating one filter criterion. Thus, based on the in silico ADME evaluation and molecular docking, it can be suggested that erythrin should be further explored to develop as a potent anti-diabetic drug.

Author contributions

Madhushree is a student and Hariprasad is the research supervisor.

Compliance with ethical standards

Conflict of interest

The authors declare that there is no conflict between them.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Madhushree M. V. Rao, Email: madhushreev.rao@gmail.com

T. P. N. Hariprasad, Email: hariprasad.tpn@gmail.com

References

  1. Arumugam G, Manjula P, Paari N. A review: anti diabetic medicinal plants used for diabetes mellitus. J Acute Dis. 2013;2(3):196–200. doi: 10.1016/S2221-6189(13)60126-2. [DOI] [Google Scholar]
  2. Bharathi A, Roopan SM, Vasavi CS, Munusami P, Gayathri GA, Gayathri M. In silico molecular docking and antidiabetic studies of dihydropyrimido [4, 5-a] acridin-2-amines. Biomed Res Int. 2014 doi: 10.1155/2014/971569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bharti SK, Kumar A, Sharma NK, Prakash O, Jaiswal SK, Krishnan S, Gupta AK, Kumar A. Tocopherol from seeds of Cucurbita pepo against diabetes: validation by in vivo experiments supported by computational docking. J Formos Med Assoc. 2013;112(11):676–690. doi: 10.1016/j.jfma.2013.08.003. [DOI] [PubMed] [Google Scholar]
  4. Bhutani R, Pathak DP, Kapoor G, Husain A, Kant R, Iqbal MA. Synthesis, molecular modelling studies and ADME prediction of benzothiazole clubbed oxadiazole-Mannich bases, and evaluation of their anti-diabetic activity through in vivo model. Bioorg Chem. 2018;77:6–15. doi: 10.1016/j.bioorg.2017.12.037. [DOI] [PubMed] [Google Scholar]
  5. Choudhary MI, Ali M, Wahab AT, Khan A, Rasheed S, Shyaula SL, Rahman AU. New antiglycation and enzyme inhibitors from Parmotrema cooperi. Sci China Chem. 2011;54(12):1926–1931. doi: 10.1007/s11426-011-4436-2. [DOI] [Google Scholar]
  6. Cui Y, Yim JH, Lee DS, Kim YC, Oh H. New diterpene furanoids from the Antarctic lichen Huea sp. Bioorg Med Chem Lett. 2012;22(24):7393–7396. doi: 10.1016/j.bmcl.2012.10.063. [DOI] [PubMed] [Google Scholar]
  7. Damián-Medina K, Salinas-Moreno Y, Milenkovic D, Figueroa-Yáñez L, Marino-Marmolejo E, Higuera-Ciapara I, Vallejo-Cardona A, Lugo-Cervantes E. In silico analysis of antidiabetic potential of phenolic compounds from blue corn (Zea mays L.) and black bean (Phaseolus vulgaris L.) Heliyon. 2020;6(3):e03632. doi: 10.1016/j.heliyon.2020.e03632. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Ganesan A, Mahesh A, Sundararaj JP, Mani K, Ponnusamy P. Antihyperglycemic and anti-oxidant activity of various fraction of Parmotrema hababianum in streptozotocin-induced diabetic rat. Bangl J Pharmacol. 2016;11(4):935–939. doi: 10.3329/bjp.v11i4.27658. [DOI] [Google Scholar]
  9. Ganugapati J, Mukkavalli S, Sahithi A. Docking studies of green tea flavonoids as insulin mimetics. Int J Comput Appl. 2011;30:48–52. [Google Scholar]
  10. Ganugapati J, Baldwa A, Lalani S. Molecular docking studies of banana flower flavonoids as insulin receptor tyrosine kinase activators as a cure for diabetes mellitus. Bioinformation. 2012;8(5):216–220. doi: 10.6026/97320630008216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Guasch L, Sala E, Mulero M, Valls C, Salvadó MJ, Pujadas G, Garcia-Vallvé S. Identification of PPARgamma partial agonists of natural origin (II): in silico prediction in natural extracts with known antidiabetic activity. PLoS One. 2013;8(2):e55889. doi: 10.1371/journal.pone.0055889. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gupta RC, Chang D, Nammi S, Bensoussan A, Bilinski K, Roufogalis BD. Interactions between antidiabetic drugs and herbs: an overview of mechanisms of action and clinical implications. Diabetol Metab Syndr. 2017;9(1):59. doi: 10.1186/s13098-017-0254-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Halama P, Van Haluwin C. Antifungal activity of lichen extracts and lichenic acids. Biocontrol. 2004;49(1):95–107. doi: 10.1023/B:BICO.0000009378.31023.ba. [DOI] [Google Scholar]
  14. Hengameh P, Rashmi S, Rajkumar HG. In vitro inhibitory activity of some lichen extracts against α-amylase enzyme. Eur J Biomed Pharm. 2016;3(5):315–318. [Google Scholar]
  15. Hidalgo ME, Fernandez E, QuilhotLissi E. Antioxidant activity of depsides and depsidones. Phytochemistry. 1994;37(6):1585–1587. doi: 10.1016/s0031-9422(00)89571-0. [DOI] [PubMed] [Google Scholar]
  16. Hu C, Jia W. Therapeutic medication against diabetes: what we have and what we expect. Adv Drug Deliv Rev. 2019;139:3–15. doi: 10.1016/j.addr.2018.11.008. [DOI] [PubMed] [Google Scholar]
  17. Huneck S. The significance of lichens and their metabolites. Naturwissenschaften. 1999;86(12):559–570. doi: 10.1007/s001140050676. [DOI] [PubMed] [Google Scholar]
  18. Jawla S, Kumar Y. Molecular docking interaction of Pinitol (ligand) with dipeptidyl peptidase 4 receptor (PDB 3C45) World Appl Sci J. 2013;24(12):1629–1634. doi: 10.5829/idosi.wasj.2013.24.12.7653. [DOI] [Google Scholar]
  19. Karthik S, Nandini KC, Kekuda PTR, Vinayaka KS, Mukunda S. Total phenol content, insecticidal and amylase inhibitory efficacy of Heterodermia leucomela(L) Ann Biol Res. 2011;2(4):38–43. [Google Scholar]
  20. Kumar R, Ramachandran U, Khanna S, Bharatam PV, Raichur S, Chakrabarti R. Synthesis, in vitro and in silico evaluation of l-tyrosine containing PPARα/γ dual agonists. Bioorg Med Chem. 2007;15(3):1547–1555. doi: 10.1016/j.bmc.2006.06.063. [DOI] [PubMed] [Google Scholar]
  21. Lee SH, Kang SM, Ko SC, Moon SH, Jeon BT, Lee DH, Jeon YJ. Octaphlorethol A: a potent α-glucosidase inhibitor isolated from Ishige foliacea shows an anti-hyperglycemic effect in mice with streptozotocin-induced diabetes. Food Funct. 2014;5(10):2602–2608. doi: 10.1039/C4FO00420E. [DOI] [PubMed] [Google Scholar]
  22. Neamati N, Hong H, Mazumder A, Wang S, Sunder S, Nicklaus MC, Pommier Y. Depsides and depsidones as inhibitors of HIV-1 integrase: discovery of novel inhibitors through 3D database searching. J Med Chem. 1997;40(6):942–951. doi: 10.1021/jm960759e. [DOI] [PubMed] [Google Scholar]
  23. Pandit R, Phadke A, Jagtap A. Antidiabetic effect of Ficus religiosa extract in streptozotocin-induced diabetic rats. J Ethnopharmacol. 2010;128(2):462–466. doi: 10.1016/j.jep.2010.01.025. [DOI] [PubMed] [Google Scholar]
  24. Patel B, Ghate M. Computational studies on structurally diverse dipeptidyl peptidase IV inhibitors: an approach for new antidiabetic drug development. Med Chem Res. 2013;22(9):4505–4521. doi: 10.1007/s00044-012-0455-6. [DOI] [Google Scholar]
  25. Prabha B, Neethu S, Krishnan SL, Sherin DR, Madhukrishnan M, Ananthakrishna R, RameshkumarKB Manojkumar TK, Jayamurthy P, Radhakrishnan KV. Antidiabetic potential of phytochemicals isolated from the stem bark of Myristica fatua Houtt. var. magnifica (Bedd.) Sinclair. Bioorg Med Chem. 2018;26(12):3461–3467. doi: 10.1016/j.bmc.2018.05.020. [DOI] [PubMed] [Google Scholar]
  26. Ranković B, Mišić M, Sukdolak S. Evaluation of antimicrobial activity of the lichens Lasalliapustulata, Parmeliasulcata, Umbilicariacrustulosa, and Umbilicariacylindrica. Microbiology. 2007;76(6):723–727. doi: 10.1134/S0026261707060112. [DOI] [Google Scholar]
  27. Rathore PK, Arathy V, Attimarad VS, Kumar P, Roy S. In-silico analysis of gymnemagenin from Gymnemasylvestre (Retz.) R. Br. with targets related to diabetes. J Theor Biol. 2016;391:95–101. doi: 10.1016/j.jtbi.2015.12.004. [DOI] [PubMed] [Google Scholar]
  28. Rozano L, Zawawi A, Redha M, Ahmad MA, Jaganath IB. Computational analysis of Gynurabicolorbioactive compounds as dipeptidyl peptidase-IV inhibitor. Adv Bioinform. 2017 doi: 10.1155/2017/5124165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Safitri A, Fatchiyah F, Sari DRT, Roosdiana A. Phytochemical screening, in vitro anti-oxidant activity, and in silico anti-diabetic activity of aqueous extracts of Ruellia tuberosa L. J Appl Pharm Sci. 2020;10(03):101–108. doi: 10.7324/JAPS.2020.103013. [DOI] [Google Scholar]
  30. Salehi B, Ata A, Anil Kumarn V, Sharopov F, Ramírez-Alarcón K, Ruiz-Ortega A, Ayatollahi SA, Fokou PVT, Kobarfard F, Zakaria ZA, IritiM Taheri Y, Martorell M, Sureda A, Setzer NW, Durazzo A, Lucarini M, Santini A, Capasso R, Ostrander EA, ur-Rahman A, Choudhary MI, Cho CW, Sharifi-Rad J. Antidiabetic potential of medicinal plants and their active components. Biomolecules. 2019;9(10):551. doi: 10.3390/biom9100551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Salin Raj P, Prathapan A, Sebastian J, Antony AK, Riya MP, Rani MRP, Biju H, Priya S, Raghu KG. Parmotrem atinctorum exhibits antioxidant, antiglycation and inhibitory activities against aldose reductase and carbohydrate digestive enzymes: an in vitro study. Nat Prod Res. 2014;28(18):1480–1484. doi: 10.1080/14786419.2014.909420. [DOI] [PubMed] [Google Scholar]
  32. Seo C, Choi YH, Ahn JS, Yim JH, Lee HK, Oh H. PTP1B inhibitory effects of tridepside and related metabolites isolated from the Antarctic lichen Umbilicaria antarctica. J Enzyme Inhib Med Chem. 2009;24(5):1133–1137. doi: 10.1080/14756360802667811. [DOI] [PubMed] [Google Scholar]
  33. Seo C, Sohn JH, Ahn JS, Yim JH, Lee HK, Oh H. Protein tyrosine phosphatase 1B inhibitory effects of depsidone and pseudodepsidone metabolites from the Antarctic lichen Stereocaulon alpinum. Bioorg Med Chem Lett. 2009;9(10):2801–2803. doi: 10.1016/j.bmcl.2009.03.108. [DOI] [PubMed] [Google Scholar]
  34. Seo C, HanYim J, Kum Lee H, Oh H. PTP1B inhibitory secondary metabolites from the Antarctic lichen Lecidellacarpathica. Mycology. 2011;2(1):18–23. doi: 10.1080/21501203.2011.554906. [DOI] [Google Scholar]
  35. Shivanna R, Parizadeh H, Garampalli RH. Screening of lichen extracts for in vitro antidiabetic activity using alpha amylase inhibitory assay. Int J Biol Pharm Res. 2015;6(5):364–367. [Google Scholar]
  36. Singh P, Singh VK, Singh AK. Molecular docking analysis of candidate compounds derived from medicinal plants with type 2 diabetes mellitus targets. Bioinformation. 2019;15(3):179–188. doi: 10.6026/97320630015179. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sivajothi V, Dakappa SS. In vitro and in silico antidiabetic activity of pyran ester derivative isolated from Tragia cannabina. Asian Pac J Trop Biomed. 2014;4:S455–S459. doi: 10.12980/APJTB.4.2014C1049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Thadani VM, Khan SN, Karunratne V, Choudhary MI (2011) α-Glucosidase inhibitors from lichens. U.S. Patent No. 7867989 dated 01/11/2011. Washington, DC: U.S. Patent and Trademark Office 2011
  39. Thadhani VM, Karunaratne V. Potential of lichen compounds as antidiabetic agents with antioxidative properties: a review. Oxid Med Cell Longev. 2017 doi: 10.1155/2017/2079697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Trinh Q, Le L. An investigation of antidiabetic activities of bioactive compounds in Euphorbia hirta Linn using molecular docking and pharmacophore. Med Chem Res. 2014;23(4):2033–2045. doi: 10.1007/s00044-013-0794-y. [DOI] [Google Scholar]
  41. Umeno A, Horie M, Murotomi K, Nakajima Y, Yoshida Y. Antioxidative and antidiabetic effects of natural polyphenols and isoflavones. Molecules. 2016;21(6):708. doi: 10.3390/molecules21060708. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Valadbeigi T. Chemical composition and enzymes inhibitory, brine shrimp larvae toxicity, antimicrobial and antioxidant activities of Caloplaca biatorina. Zahedan J Res Med Sci. 2016;18(11):e4267. doi: 10.17795/zjrms-4267. [DOI] [Google Scholar]
  43. Valadbeigi T, Shaddel M. Amylase inhibitory activity of some macrolichens in Mazandaran province, Iran. Physiol Pharmacol. 2016;20(4):215–219. [Google Scholar]
  44. Verma N, Behera BC, Om Sharma B. Glucosidase inhibitory and radical scavenging properties of lichen metabolites salazinic acid, sekikaic acid and usnic acid. Hacettepe J Biol Chem. 2012;40(1):7–21. [Google Scholar]
  45. Vinayaka KS, Karthik S, Nandini KC, Kekuda PTR. Amylase inhibitory activity of some macrolichens of Western Ghats, Karnataka, India. Indian J Nov Drug Deliv. 2013;5(4):225–228. [Google Scholar]
  46. Vo THN, Tran N, Nguyen D, Le L. An in silico study on antidiabetic activity of bioactive compounds in Euphorbia thymifolia Linn. SpringerPlus. 2016;5(1):1359. doi: 10.1186/s40064-016-2631-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wade RC, Goodford PJ. The role of hydrogen-bonds in drug binding. Prog Clin Biol Res. 1989;289:433–444. [PubMed] [Google Scholar]
  48. Zhang Y, Shi J, Zhao Y, Cui H, Ca C, Liu S. An investigation of the anti-diabetic effects of an extract from Cladonia humilis. Pak J Pharm Sci. 2012;25(3):509–512. [PubMed] [Google Scholar]

Articles from In Silico Pharmacology are provided here courtesy of Springer

RESOURCES