In silico exploration of Elaeocarpus ganitrus extract phytochemicals on STAT3, to assess their anticancer potential

Cancer is a cluster of diseases and STAT3 protein has significant roles in all types of cancer. Signal transducer and activator of transcription (STAT), belong to the family of cytoplasmic transcription factors, activate and transduce extracellular growth factor, and also affect cytokine signals and affect gene transcriptional events. STAT3 mutant intrinsically alone is enough to instigate oncogenic transformation, and tumorigenesis^[1−3]. A survey of the current literature reveals that STATs have transactivated domains and play a significant role in cancer migration and invasion. Hampering of c-Src kinase activity or downregulation of STAT3 signaling stimulates apoptosis^[4]. The study of chemical interactions between STAT3 receptor and phytochemicals assist in drug designing and hence in cancer therapy^[5]. There are a variety of phytochemicals that have a high propensity to modulate directly or indirectly the STAT3 signaling pathway. Triterpenoids like betulinic acid, polyphenols curcuminoids, plumbagin a naphthoquinone, diosgenin a steroid, hydroxycinnamic acid, and thymoquinone are the phytochemicals that suppress STAT3 expression^[6]. Many plant-derived phytochemicals manifest high anticancer activity and lead researchers to adopt integrated multifaceted research techniques^[7−11]. Though, Elaeocarpus ganitrus Roxb. (also known as Rudraksha) constitutively placed in Ayurvedic system, also has anticancer potential^[12]. Recently its silver nanoparticle has been assessed for anticancer and antiproliferative activities^[13]. Its impact on STAT is yet to be explored. In the last few decades, phytochemical composition of Elaeocarpus genus has been extensively investigated. Phytochemicals of various extracts of different parts of the plant showed the presence of alkaloids, flavonoids, carbohydrates, glycosides, proteins, quinine, coumarins, tannins, minerals, vitamins, saponins, phenolic compounds, and fixed oils in a high concentration, thus adding to its medicinal value^[14]. The pharmacological screening of metabolites like polyphenols, alkaloids, terpenoids and flavonoids have been explored to demonstrate cancer pathways to ascertain possible mechanism^[15−18]. As stated in the literature, the beads and the bark of the plants have been extensively studied while the leaves of the E. ganitrus have not been studied for their anticancer efficiency. Besides, leaves of the plants were shown to have good antioxidant potential^[19,20]. The emphasis of the study is to identify phytochemicals retrieved from Elaeocarpus ganitrus leaf research data and GC-MS profiling. To accentuate, the binding role of Elaeocarpus ganitrus phytochemicals with STAT3 receptor, their ADME properties and pharmacokinetic studies were investigated.

The chemicals and solvents used in the extraction and phytochemical analysis were of analytical grade, sourced from Sigma-Aldrich. MTT (3-[4,5-dimethylthiazol-2-1yl]-2,5 diphenyl tetrazolium bromide) was also procured from Sigma-Aldrich. HeLa cells were obtained from the National Centre for Cell Sciences (NCCS), Pune, India. Fetal bovine serum and Dulbecco's Modified Eagle's Media were acquired from Gibco-life technologies.

Procurement of plant material

Fresh leaves of E. ganitrus were purchased from Patanjali Herbal Garden Nursery in Panchayanpur, Uttarakhand, India. Authentication of the Elaeocarpus ganitrus was conducted by the Department of Botany, Jamia Hamdard, New Delhi, India, and the voucher specimen was deposited at the University.

Method to prepare extracts

Leaves of E. ganitrus were carefully washed, air-dried for ten days, and ground to a fine powder. A sample of 1,000 grams of powder was exhaustively extracted three times with 100% methanol (10 times weight/volume) at room temperature for 72 h using a soxhlet apparatus. The resulting crude methanol extract was fractionated successively with solvents in increasing polarity order: heptane, chloroform, ethyl acetate, methanol, and water. The residue was air-dried and utilized for the subsequent solvents. The fractions obtained from each solvent were filtered, dried under vacuum using a rotary evaporator, and stored at 40 °C until use^[21].

Qualitative phytochemical analysis

The presence or absence of phytochemicals such as terpenoids, steroids, saponins, flavonoids, glycosides, tannins, and phenols in the chloroform and methanol extracts of E. ganitrus leaves was determined following the standard methodology^[22].

Cellular studies on HeLa cancer cell line

The HeLa cell line was stored in Dulbecco's Modified Eagle's Medium which is rich in 10% Fetal Bovine Serum, 1% antibiotic solution, 25 mM sodium bicarbonate, and 10 mM HEPES in a 5% CO₂ humidified atmosphere at 37 °C in an air jacketed incubator. The stock culture was perpetuated in the exponentially growing phase by passaging as, monolayer culture with 0.02% EDTA. Dislodged cells suspended in complete medium were routinely reseeded.

Cytotoxicity assay/MTT assay

The cytotoxic effects of the various fractions of E. ganitrus leaf on the HeLa cancer cell line were evaluated using the MTT assay. Cells were seeded overnight, and exposed to different concentrations of the prepared fractions (ranging from 50 to 200 μg/ml), and incubated for 48 h. After treatment, cells were incubated with MTT solution and the formazan crystals were solubilized and the absorbance was read at 570 nm^[23].

In silico investigation of phytochemicals obtained from leaf extracts

Binding energies of phytochemicals retrieved from plant leave extract with STAT3 were calculated by using software InstaDock for molecular docking. Discovery Studio Visualizer, and PyMOL, were used to visualize the chemical interactions of ligands and proteins. SWISS-ADME tool and ProTox-II were used for pharmacokinetic profiling studies. The X-ray crystal structure of STAT3 (PDB ID: 6NJS) was downloaded from Protein Data Bank (PDB). All co-crystallized hetero atoms and attached water molecules and co-crystallized ligands, were eliminated from the original coordinates. The Polar hydrogen atoms were inculcated, the residue structures having lower occupancy were removed, and the incomplete side chains were then substituted by using ADT. Three-dimensional structures of phytocompounds were sketched using Chem3D.

Drug-likeness properties prognosis

Determination of the analogous behavior to the drug of phytoligands with the help of cheminformatics was done using online tool SwissADME developed by the Molecular Modelling Group, Swiss Institute of Bioinformatics^[24]. The computation of pharmacokinetics and physicochemical molecular properties help medicinal chemists in their routine drug discovery processes. Significant basic molecular information can be excavated from the chemical structure. The methods were preferred over other methods because of the speed, but also for the ease of interpretating results by fingerprinting method to enable researchers move through translation to medicinal chemistry and in molecular designing^[25].

Molecular docking-based virtual screening

The rationale behind molecular docking is to steer medicinal chemists for translational research. The affinity of a molecule to the receptor changes with small structural changes in the molecule^[26]. For molecular docking, STAT3 core complex PDB id (PMID: 31715132) was remodeled to ascertain binding energies with the best conformational poses of Elaeocarpus ganitrus leaves phytoligands. The InstaDock software is used to dock phytoligands with blind search space having a grid size of 110, 70, and 108 Å for X, Y, Z coordinates, correspondingly. The center of the grid was confined to X: 63.09, Y: 14.98, and Z: −76.91 axis, which covers all the heavy atoms embedded in the protein. The conformational site selected was so that the movement of the ligands was free to probe their best binding coordinates. Default docking specifications were employed to calculate various parameters. All the docking conformational poses were generated using PyMOL, a molecular visualization system and Discovery Studio Predictor.

Pharmacokinetic and toxicity prediction

Physicochemical parameters, water solubility, lipophilicity, pharmacokinetics, and drug-likeness were elicited from SwissADME. To retrieve the toxicological profile of the phytoligands ProTox-II servers were employed^[27]. Early estimation of the Absorption, Distribution, Metabolism, Excretion and Toxicity abbreviated as ADMET imperative to ascribe the pharmacodynamics success of the lead phytoligands. (SMILES) strings to encode chemical structures were imported from PubChem, open chemistry database and implemented in SWISS-ADME tool^[24] to auspicate lipophilicity to show hydrophobic and hydrophilic nature, water solubility, necessary for absorption across membranes, and drug-likeness rules to assess metabolic profiles. Toxicology prediction of phytoligands is a crucial and fundamental aspect in the drug discovery process. ProTox-II is used to estimate computational toxicity, to accelerate the course to drug discovery, compute animal toxicity, and also help to attenuate animal experiments. In the PROTOX-II web server, toxicity classes are designated into four segments. Category I comprised of chemical entities with LD50 (LD = lethal Dose) values (LD50 ≤ 5) mg/kg, Category II comprised of compounds with LD50 values (5 < LD50 ≤ 50) mg/kg, Category III comprised of chemical entities having LD50 values (50 < LD50 ≤ 300) mg/kg, Category IV comprised of compounds which have LD50 values (300 < LD50 ≤ 2,000) mg/kg, Category V comprised of compounds with LD50 values (2,000 < LD50 ≤ 5,000) mg/kg and Category VI comprised of compounds showing LD50 values (LD50 > 5,000) mg/kg^[28]. Category I and II manifested high toxicity, Category III and IV are comparatively less toxic and Category V and VI are considered to be non-toxic.

The solvent extraction technique is usually employed to prepare extracts from plant materials attributable to its convenience to operate. The importance lies in that a large amount of plant material can be extracted with minimal solvent^[26]. Fresh leaves of Elaeocarpus ganitrus were purchased from Patanjali Herbal Garden Site Nursery located in Panchayanpur, Uttarakhand 249405, India. The confirmation of the authenticity of the Elaeocarpus ganitrus was done by the Department of Botany, Jamia Hamdard, New Delhi, India, and the leaf specimens deposited in the University. The crude methanol extract was unintermittedly fractionated in the solvents heptane, chloroform, ethyl acetate, methanol, and water according to their increasing polarity^[16]. The anticancer activity of extracts was analyzed on the basis of their IC₅₀ values. Cancerous HeLa cell line when treated with E. ganitrus leaf extracts exhibited a substantial inhibition of cells. The half maximal inhibitory concentration of chloroform and methanol extracts of E. ganitrus was (IC₅₀ = 304.39 μg/ml) and (IC₅₀ = 308.59 μg/ml) respectively followed by water (IC₅₀ = 340.14 μg/ml), ethyl acetate (IC₅₀ = 350.72 μg/ml) and heptane (IC₅₀ = 381.76 μg/ml) extracts (Fig. 1a−e & Fig. 2). The qualitative investigation using standard methodology^[22] of chloroform and methanol fractions of E. ganitrus leaves disinterred the presence of major phytochemicals namely steroids, saponins, terpenoids, tannins, phenols, glycosides and flavonoids Table 1. GC-MS analysis of the chloroform and methanolic fractions was done based on their lowest half maximal inhibitory concentration to get a complete profiling of the plant compounds. The peaks in the total ion current (TIC) chromatogram of GC-MS profile of the phytoligands commensurate with the spectrum of known chemical databases stockpiled in the GC-MS library. The gas chromatogram depicts the relative concentrations of different phytoligands getting eluted according to the retention time. The heights of the peak represent the comparative concentrations of the compounds present in the plant appear as peaks at different m/z ratios. The components present with their retention time, molecular formula, molecular weight and concentration (peak area %) are provided in Tables 2 & 3 showing the presence of 56 and 50 bioactive phytochemicals in the chloroform and methanol extracts respectively. Of 106 phytoligands obtained from chloroform and methanol extracts of E. ganitrus leaves, 81 phytoligands were identified has having the best drug-like properties following Lipinski's rule of five. Lipinski's rule states that molecular properties, physical or chemical of a compound are significant for a drug's pharmacokinetics behavior inside a biological system. The drug molecules that go along with the RO5 have fewer attrition rates when undergoing clinical trials. The cheminformatics study to identify potential chemical entities having propensity for predefined biological targets is called virtual screening^[28]. To endeavor in vitro experiments time diminution, molecular docking-based virtual screening of 81 selected compounds with two reference inhibitors having substantial binding energies with 6NJS were preferred for further analysis. The STAT3 has dual nature as an oncogene or as a tumor suppressor during cancer progression. It has a SH₂ domain, linker domain, DNA binding domain, and all-alpha domain. The total energy of binding, Vander Waals forces, hydrogen bonding, electrostatic attraction, desolvation, and also a number of rotatable bonds present in the phytoligand, contribute to observe the free energy of binding of phytoligands with the receptor. Twenty-six (EG-1 to EG-26) compounds were selected as having appreciable binding affinities towards the 6NJS receptor (Table 4).

Figure 1. 

Effects of (a) heptane, (b) chloroform, (c) methanol, (d) ethyl acetate and (e) water fractions of E. ganitrus leaves on the human cancer cell lines HeLa using MTT assay.

Figure 2. 

IC₅₀ values of different extracts of E. ganitrus leaves against human cancer cell lines HeLa.

The absorption of drugs by the body is related to their pharmacokinetic properties and also cellular toxicity. The potency of the drug depends mostly on the pharmacokinetic parameters because ADME processes command the rate and extent of absorption when an administered dose of a drug approaches to its action site. Hence, in silico pharmacokinetic profile of filtered compounds was surveyed to gather the putative bioavailability data for receptor 6NJS. The cumulative findings for pharmacokinetics profiling, bioavailability data, drug-likeness properties and drug friendliness and toxicity effects of selected 26 phytoligands with known inhibitors (Plumbagin and Sanguinarine) are given in Tables 5−10. The prediction revealed that the six molecules (EG-9, EG-12, EG-13, EG-16, and EG-26) can be lead compounds for new drug candidates for anti-cancer phytomedicine. The Half maximal Inhibitory concentration of EG-13 was (IC₅₀ = 254.29 µg/ml) further support our results (Fig. 3).

Figure 3. 

IC₅₀ values of EG-13 phytochemical of E. ganitrus leaves against human cancer cell lines HeLa.

In Table 5, for pharmacokinetics prognostication, the gastrointestinal (GI) absorption rate was fetched for all preferred six phytoligands and both reference drugs. The blood-brain permeability was seen as positive for all the six phytoligands and both reference drugs. The prediction of bioavailability (Table 6) demonstrated that similar bioavailability scores were observed for all the filtered six phytoligands (0.55) like reference drugs. The water solubility data showed all the six compounds and plumbagin are soluble while Sanguinarine is poorly soluble. For drug-likeness prediction (Table 7), all the six compounds and both known inhibitors were obtained suitable for the Lipinski rule as zero violation. For Ghose, Veber, and Egan filter 0 violation was obtained for all the six phytoligands and both inhibitors. In the case of medicinal chemistry friendliness prediction (Table 8), the PAINS structural alert obtained 0 violations for all the six phytoligands and sanguinarine while two alerts for plumbagin. Table 9 shows EG-9 belongs to the non-toxic class VI, EG-15, and EG-16 also belong to the non-toxic class V, EG-12, EG-13, EG-26 and Sanguinarine belongs to the less toxic class IV while plumbagin belongs to the high-toxic class II. The bioavailability radar (Fig. 3) for phytoligands depicting bioavailability prognostic showed that all six phytoligands were found within the data range of lipophility nature (−0.7 < XLOGP3 < +5.0), molecule size (150 g/mol < MW < 500g/mol), polarity (20 Ų < TPSA < 130Ų), insolubility [−6 < LogS (ESOL) < 0], insaturation (0.25 < Fraction Csp3 < 1) and flexible bonds (0 < Num. rotatable bonds < 9) and colored part of radar while known inhibitors plumbagin and sanguinarine does not fit the bioavailability radar (Table 10). As mentioned in Table 5, all the phytoligands and reference compounds have higher gastrointestinal (GI) absorption rates, therefore they can instantly be absorbed by the human intestine. All phytoligands have the ability to pass the blood brain barrier (BBB permeant) and values for the aqueous solubility (log S) of the phytochemicals fall in the recommended range that is −1 to −5^[29], thus, have improved absorption and distribution properties. The bioavailability scores were identical for all six molecules, standing at 0.55, similar to the reference drugs. In drug-likeness prediction, none of the six compounds and both known inhibitors violated the Lipinski rule, Ghose, Veber, and Egan filters. Regarding medicinal chemistry friendliness, the PAINS structural alert identified zero violations for all six phytoligands and Sanguinarine, whereas Plumbagin had two alerts. Table 9 revealed that EG-9 belonged to the non-toxic class VI, while EG-15 and EG-16 were in harmless class V. Other compounds EG-12, EG-13, EG-26 and sanguinarine was from less harmful class IV which could be modified to a non-toxic class during the lead optimization stage of drug discovery^[30] while selected standard plumbagin showed high toxic class II. Drug-induced hepatotoxicity often lead to abrupt liver failure and drug rejections^[31]. Drug-induced liver injury might be long-term or occur only once. Obviously, the selected compounds and standards are non-hepatotoxic. The bioavailability radar (Fig. 4) depicted that all six phytoligands were within the data range for oral bioavailability prediction. Conversely, standards plumbagin and sanguinarine did not fit within the bioavailability radar. The pink area shown in the radar corresponds to the most promising zone for all the bioavailability properties. In Table 10, all the phytochemicals satisfied 150 g/mol and 500 g/mol criteria for (SIZE) of good drug candidates. The polarity (POLAR) was observed with the Total Polarity Surface Area (TPSA) and all the phytochemicals show good TPSA values. Besides, the flexibility (FLEX) property evaluated by the number of rotatable bonds falls within the recommended range. Lipophilicity (LIPO) and insolubility (INSOLU) were evaluated and come in the range The Unsaturation (INSATU) was calculated using Fraction Csp3 falls within a recommended range of 0.25 < Fraction Csp3 < 1) for all phytoligands. However, Plumbagin and Sanguinarine exhibit lower values (0.09 and 0.15, respectively).

Figure 4. 

Bioavailability radar (pink area exhibits optimal range of particular property) for leading phytocompounds molecules. LIPO = lipophilicity as XLOGP3, SIZE = size as molecular weight, POLAR = polarity as TPSA (topological polar surface area), INSOLU = insolubility in water by log S scale, INSATU = insaturation as per fraction of carbons in the sp3 hybridization, and FLEX = flexibility as per rotatable bonds.

2D and 3D interactions of the five phytoligands (EG-9, EG-12, EG-13, EG-15, EG-16 and EG-26) with 6njs are shown in Table 11. EG-9 divulged two assenting hydrogen bond interactions at the active site having amino acids of Glu96 and Lys97. In additon to that a non-classical C-H bond Vander Waals interaction was also noticed at the active site involving Arg93 residue and alkyl and pi-alkyl interactions were observed at Leu525 and Trp501 respectively. In EG-12 a conventional hydrogen bond interaction was observed at Asn538, a pi-pi T-shaped, two alkyl and a pi-alkyl interactions were observed at Tyr539, Ile522, Trp501 and Leu525 respectively. EG-13 showed one favorable hydrogen bond interaction and two hydrophobic alkyl interactions at the active site with the residue of Glu96, Leu95 and Lys97 respectively. EG-15 showed two alkyl and two pi-alkyl interactions at the active site of the residues of Leu95, Ile522, Trp501 and Tyr539 respectively. In EG-16 two conventional hydrogen bonds were observed at Leu731 and Thr716. EG-26 formed three favorable hydrogen bonds with Asp369, Asp370 and Asp371 at the active site of the receptor. Plumbagin showed a conventional hydrogen bond interaction, a pi-pi T-shaped and a carbon-hydrogen bond interaction at Tyr539, Trp501 and Ser540 respectively. Sanguinarine showed a carbon hydrogen bond, a pi-sigma, a alkyl, and a pi-alkyl interaction at the site of Glu696, Leu731, Pro769 and Pro695 respectively (Table 11). Previously it has been shown that residue at 97 could have amprospective ubiquitin acceptor position in STAT3 NH₂ terminal domain, suggesting lysine amino acid may have a significant role and location in a sumolation/ubiquitination consensus sequence^[32]. The majority of phytoligand interactions exist in the Linker domain and Transactivation domain of the STAT3.

Table 11.  2D and 3D binding interactions between the receptor 6NJS and molecules.

Phyto-ligands	2D- Binding interaction	3D- Binding interaction
EG-9 (-6.8)
EG-12 (-6.5)
EG-13 (-6.5)
EG-15 (-6.5)
EG-16 (-6.4)
EG-26 (-5.7)
Plumbagin
Sanguinarine

All the six compounds (EG-9, EG-12, EG-13, EG-15, EG-16 and EG-26) significantly bind with STAT3. The phytochemicals epitomized good in silico results as reflected by their promising binding affinity, considerable inhibitory constant with optimum protein-ligand stabilization energy. Consecutively, binding signifies that phytoligands interact with STAT3 by the NH₂ terminal and boosts its transcriptional activity and interferes with the cellular proliferation process and apoptosis^[32]. Bioavailability radar and toxicological profiles of the preferred phytoligands revealed that these compounds compel to have ample drug likeliness properties. Moreover, EG-9, EG-13, EG-15, EG-16 and EG-26 have not been explored for their anticancer potential and can be derivatized or have the probability of being used as lead compounds.

The authors confirm contribution to the paper as follows: study design and draft manuscript preparation (equal): Mehnaj, Bhat AR, Athar F; supervision: Athar F; experimentation and writing of manuscript: Mehnaj; characterization and editing: Bhat AR. All authors reviewed the results and approved the final version of the manuscript.

The supplementary data will be made available by the authors to all upon reasonable request.

Miss Mehnaj is grateful to UGC for obtaining the non-NET fellowship allowing completion of this work.