In silico and in vitro approaches to evaluate the bioactivity of Cassia auriculata L extracts

Abstract

The present study is an attempt to evaluate the in vitro anti‐inflammatory and in silico anticancer potentials of the plant Cassia auriculata (CA). The aerial parts of CA were subjected to solvent extraction, and the extracts were fractionised by gas chromatography and mass spectrometry analysis for its phytochemical content. The antiinflammatory activity of the extracts were confirmed by the IC₅₀ value of 125.02 µg/ml for red blood cell membrane stabilisation and 195.7 µg/ml for inhibition of protein denaturation activity. The interaction of bioactive compounds of CA ethanol extract with target protein was predicted through molecular docking studies, serine/threonine–protein kinase B (AKT1), responsible for development and progression of lung cancer using AutoDock tools. Extensive studies have been carried out on a range of kinase inhibitors targeting Akt, but obtaining promising results is a challenge yet due to its toxicity and resistance issues. Yohimbine, undecanoic acid 10‐methyl‐ethyl ester and chrysin significantly bind to the target protein with least binding energy. Hence, the present paper establishes the anti‐inflammatory and anticancer capacities of CA ethanol extract as an alternative to the existing therapeutic approach to inflammation and cancer through a systematic in vitro and in silico approaches supplementing the findings.

1 Introduction

The relation between antioxidants and degenerative diseases associated with inflammation is the main topic of focus by many researchers nowadays [1]. Inflammation and oxidative stresses are interrelated processes and both are responsible for the pathogenesis of various degenerative diseases in humans such as cancer, ageing, inflammatory and rheumatoid arthritis. So, it is necessary to identify anti‐inflammatory agents from exogenous sources to reduce reactive oxygen species (ROS) action.

Lung cancer is the most common amongst human cancers and is the leading cause of cancer deaths globally. Even though advancements in treatments with new drugs and targeted therapies, the disease is often detected during the advanced stages, where the five‐year survival rate is close to 1% [2]. Akt or protein kinase B (PKB) plays a central role in cell signalling pathways and controls cell survival and death [3], the activation of which, in turn, associated with resistance to apoptosis and cancer cell survival, growth, migration and energy metabolism [4, 5, 6]. The anti‐apoptotic nature of Akt is responsible for its transforming ability and resistance of the cancer cells to the chemotherapy and radiotherapy. From the literature, it is inferred that Akt confers a growth advantage to the cancer cells and it maybe essential in controlling their growth, survival and migration. Serine/threonine kinase 1 (gene encoding for protein serine/threonine kinase) (AKT1) is one of the three isoforms of Akt and it is expressed ubiquitously in all the tissue types.

Lung cancer has been frequently linked with overactivated protein Akt/PKB. There are several evidence points to the role of deregulated Akt in the progression of lung cancer [7, 8]. Also, recent findings demonstrated that the cancer cells defend themselves from therapeutic treatment through the activation of pro‐survival signals, mainly the Akt pathway. The role of AKT signalling in lung tumorigenesis includes the hyperactivation of Akt protein in lung cancer [9]. Recent studies discovered Akt1 is activated via somatic mutation in the PH domain, which leads to the constitutive activation of Akt1 in human lung cancer patients.

During the localisation event at the plasma membrane, the Akt gets phosphorylated, which induces protein synthesis and growth of cells through the activation of mammalian target of rapamycin. From these findings, Akt is observed to be an attractive target to overcome the issue of lung cancer [10]. Therefore, the inhibition of Akt by plant‐derived molecules is of immense therapeutic significance for the development of anticancer drugs.

In recent days, more emphasis has been given for identifying plant‐derived compounds to effectively treat life‐threatening diseases such as cancer, which is caused due to abnormalities in the DNA sequence of cell [11]. Mutations in the DNA sequence alter a cell from its normal phenotype to cancerous type; identifying such mutated genes is the central aim of cancer research. The advancement in the field of cheminformatics now led the way to identify the therapeutic properties of chemical compounds deposited in chemical databases. Computer‐based docking techniques play an important role in drug design and elucidating the mechanism of action; these are when used before experimental screening, reduce labour and cost‐effective for the development of effective medicinal compounds [12, 13].

The flower extract of the plant is used to treat urinary discharges, diabetes, nocturnal emissions and throat irritation. The leaves of the plant are reported to possess anthelmintic, anti‐ulcer and used against skin diseases and leprosy [14]. The anticancer activity of Cassia auriculata (CA) was reported in various cancer cell models [15], but the exact mechanism of action on anticancer activity has not been elucidated in human lung cancer cells.

The present study attempts to identify key bioactive compounds present in the selected plant source by employing the gas chromatography and mass spectrometry (GC–MS) technique and also in vitro experiments were performed to evaluate the anti‐inflammatory potential of this selected drug. Although there have been extensive studies carried out on AKT1, only a few have addressed on its binding interaction with phytocomponents. Therefore, the present paper is an effort to determine the efficiency of interaction between the phytochemical constituents of the ethanolic extract of CA with the AKT1 target protein through molecular docking studies.

2 Materials and methods

2.1 Collection, extraction and concentration of plant material

The aerial parts of the plant CA were collected from the outer rural areas of Chennai, Tamil Nadu, India. The plant was authenticated by a botanist Dr. P.T. Devarajan from Presidency College, Chennai, India. The plant specimen was kept in the herbarium of botany department, Presidency College. About 100 g of powdered material was subject to cold percolation process with the solvents of increasing polarity, namely n ‐hexane, di‐chloro methane (DCM), ethyl acetate and ethanol. The extracts were concentrated using a rotary evaporator at 40°C under reduced pressure and stored at 4°C until further use.

2.2 GC–MS analysis

All the four extracts were subjected to GC–MS analysis to enumerate its phytochemical compounds using the GC–MS instrument (JEOL GCMATE II GC–MS, Japan) equipped with a secondary electron multiplier. The column (HP5) was fused with silica 50 m × 0.25 mm internal diameter (diameter of the GC‐MS column) (ID). The initial temperature of the instrument was maintained at 110°C for 2 min. The oven temperature was increased to 280°C at a rate of 5°C/min and maintained for 9 min at the end. Helium was used as carrier gas at a continuous flow rate of 1 ml/min and the injection port temperature was ensured at 250°C. The ionisation voltage was 70 eV. The samples were injected in split mode at 10:1. Mass spectral scan range was at the rate of 45–450 m/z. Using computer searches on National Institute of Standards and Technology Ver.2.1 MS data library [16, 17] comparing the spectrum obtained through GC–MS, compounds present in the plant extracts were identified.

In addition to anti‐inflammatory activity, the major active compounds of CA ethanolic extract were subjected to computerised docking techniques to predict the binding orientation of identified natural compounds with their protein targets.

2.3 Determination of in vitro anti‐inflammatory activity

2.3.1 Inhibition of BSA denaturation

The reaction mixture contains test extracts of varying concentrations 1.8 ml (50, 100, 200 and 400 µg/ml) and 0.2 ml of 1% aqueous bovine serum albumin. The pleckstrin homology (PH) of the solution was adjusted to 6.5 and the reaction mixture was incubated at 37°C for 20 min. Then heated to 57°C for 15 min. The turbidity was measured at 660 nm. Control was maintained parallel without extracts. The test was done in triplicate and the percentage inhibition of bovine serum albumin (BSA) denaturation was calculated as percentage inhibition = (Abs_control –Abs_sample) × 100/Abs_control.

2.3.2 Human red blood cell (HRBC) membrane‐stabilisation method

About 2 ml of fresh whole human blood was collected and mixed with an equal volume of sterilised Alsevers solution (2% dextrose, 0.8% sodium citrate, 0.5% citric acid and 0.42% sodium chloride) and centrifuged at 3000 rpm. The cells were washed with an equal volume of isosaline. Then using normal saline, 10% v/v suspension was made and stored at 4°C. Different concentrations (50, 100, 200 and 400 µg/ml) of CA extract were made with normal saline. Aspirin was used as standard and distilled water instead of hyposaline (to get 100% haemolysis) was kept as control. The reaction mixture was incubated at 37°C for 30 min and centrifuged at 3000 rpm for 20 min. The haemoglobin content of the supernatant was estimated spectrophotometrically (UV‐2700, Shimadzu, Singapore) at 560 nm. The percentage stabilisation of HRBC membrane was calculated by the formula mentioned in Section 2.3.1.

2.4 Determination of in silico anticancer activity

2.4.1 Preparation of target protein and ligand structure

In the present paper, the crystal structure of AKT1 [Protein Data Bank (PDB) ID: 4EJN] from Homo sapiens was retrieved from PDB research collaboratory for structural bioinformatics (RCSB) (Fig. 1). The structures of ligands were retrieved from PubChem database and drawn using advanced chemistry development Inc. (ACD)/ChemSketch, saved in. mol format and converted to. pdb format using Open Babel molecular converter tool. Three‐dimensional (3D) optimisation of ligand structures was done using ChemSketch software. The compounds under investigation are listed in Fig. 2. The hydrogen bond interaction between the compound and drug target was visualised using Discovery Studio Visualiser version 3.1 (Accelrys, San Diego, USA). The optimised ligands were docked with target protein AKT1 using ‘ligand fit’ model in AutoDock 4.2 [18, 19] and energy interface between protein and ligand were calculated.

Fig. 1.

X‐ray crystallographic structure of AKT1 visualised using RasMol (PDB code: 4EJN)

Fig. 2.

Name, chemical structure and molecular properties of the phytoligands of CA ethanolic extract (a) Hydrogen acceptor, (b) Hydrogen donor

2.4.2 Protein–ligand docking using AutoDock tools

The searching grid extended over the chosen target protein. Inclusion of polar hydrogen, assigning Kollman charges and addition of atomic solvation factors were done on ligand moieties. The 3D structure of the target protein was docked with different phytochemical compounds (Table 1) of ethanolic extract of CA derived from GC–MS analysis using AutoDock V.4.2 software. A grid spacing of 0.375 Å between the points was used and the search was carried out by the Lamarckian genetic algorithm. Various docked conformations were obtained after several runs of AutoDock and it was used to examine the predicted docking energy. From the docked structures, the best ligand–receptor assemblies were selected based on the lowest energy and marginal solvent availability of the ligand [20]. The docking results were visualised by Accelrys Visualiser discovery studio tool (Accelrys).

Table 1.

Bio active compounds identified from aerial parts of CA ethanolic extract by GC–MS analysis

Sl. no	Name of chemical compound	Molecular formula	Retention time minimum	Percentage content
1.	oleic acid, methyl ester	C19 H36 O2	18.85	14.4
2.	thujopsene‐(12)	C15 H24	16.22	2.7
3.	undecanoic acid, 10‐methyl‐, methyl ester	C13 H26 O2	17.2	23.8
4.	chrysin	C15 H10 O4	17.9	16.9
5.	oleic acid	C18 H34 O2	19.63	10.1
6.	oxiraneundeanoic, 3‐phenyl‐, methyl ester, trans‐	C10 H10 O2	20.58	7.5
7.	tricosane 2,4‐dione	C23 H44 O2	21.72	5.9
8.	yohimbine	C21 H26 N2 O3	21.9	21.6

2.5 Statistical analysis

All the data represented were with mean±standard deviations (SDs) of three parallel readings. The differences among the groups in all studied parameters were analysed by one‐way analysis of variance. Dunnett's multiple comparisons test was used to compare each treatment group with a single control. All the statistical analysis were done by using GraphPad Prism software version 8.0 (San Diego, USA). P values ˂0.05 were considered statistically significant.

3 Results and discussion

3.1 GC–MS analysis

The GC–MS analysis of all the four extracts of CA aerial parts such as n ‐hexane, DCM, ethylacetate (data not shown) and ethanol extract was done to find out the bioactive compounds. The ethanolic extract exhibited eight peaks on chromatogram (Fig. 3). The identified compounds of the extract were grouped into long‐chain fatty acids, alcohols, hydrocarbons, phenolics and esters (Table 1).

Fig. 3.

GC–MS analysis of CA ethanol extract

(A) GC–MS chromatogram of CA ethanol extract analysed on GC system and it showed the presence of eight compounds, (B) Names of chemical constituents are shown here. (a) Thujopsene‐(12), (b) Undecanoic acid, 10‐methyl‐, methyl ester, (c) Chrysin, (d) Oleic acid, methyl ester, (e) Oleic acid, (f) Oxiraneundeanoic, 3‐phenyl‐, methyl ester, trans‐, (g) Yohimbine, (h) Tricosane 2,4‐dione

Soundharajan and Ponnusamy [21] reported 21 compounds from the methanolic extract of CA aerial parts. In a previously reported paper, the ethanolic flower extract of CA showed eight compounds on GC–MS analysis [22] and the variation in the active components from the previous studies might be due to different extraction procedures and other experimental conditions opted for the present paper. The bioactive components from the present paper such as oxiraneundecanoic, 3‐phenyl–methyl ester trans, which are considered as a phenolic analogue of cinnamic acid, a potential anti‐tumour agent [23]. Oleic acid, oleic acid methyl ester (long‐chain fatty acids) and thujopsene, a sesquiterpene are also measured as a potential anticancer agents [24]. The compounds such as chrysin, a flavonoid and thujopsene found in the ethanolic extract might be responsible for the anti‐inflammatory potential of CA since, and these are reported to be effective against inflammation [25, 26].

3.2 Anti‐inflammatory activity

Analysing new chemical compounds using animal models pose certain ethical challenges and the absence of rationale to use those animals for pharmacological research is a major disadvantage. Therefore, if other suitable methods or alternatives are available that could be investigated. We have selected here the bioassay of protein denaturation and membrane‐stabilisation potential as an in vitro evaluation of anti‐inflammatory property of the selected plant material.

3.2.1 Inhibition of BSA denaturation

Maximum inhibition of heat‐induced BSA denaturation was noted in the ethanol extract of CA (IC₅₀ : 195.7 ± 5.7 µg/ml) followed by ethyl acetate extract (IC50: 200.59 ± 8.5 µg/ml) (Fig. 4). From the results obtained, it is inferred that both the extracts showed promising activity and it is significantly comparable with the standard non‐steroidal anti‐inflammatory drug aspirin (IC₅₀ : 321.5 µg/ml). The results of the present paper revealed that the extracts can inhibit protein denaturation induced by heat in a dose‐dependent manner. The presented results are in agreement with this paper already reported by Uppin et al. [27] and Mali et al. [28] who have confirmed a significant anti‐inflammatory effect through in vitro model and also demonstrated the same using carrageenan‐induced rat paw oedema methanolic extracts of CA leaves, respectively.

Fig. 4.

Anti‐inflammatory activity based on inhibition of BSA denaturation assay. Values are means±SD, P˂0.05

3.2.2 Membrane‐stabilisation potential

During the inflammatory response, the lysosomal constituents such as proteases and bactericidal enzymes are released from the activated neutrophil, which, in turn, cause supplementary tissue damage and inflammation. Since the human erythrocyte membrane mimics the lysosomal membrane components, the inhibition of hypotonicity‐induced haemolysis was used to examine the anti‐inflammatory potential of plant extracts under study. Results represented in Fig. 5 indicated that CA ethanol extract exhibited the highest inhibition of haemolysis (IC₅₀ : 125 ± 3.57 µg/ml) followed by ethyl acetate extract (IC₅₀ : 200.05 ± 5.06 µg/ml). The results of RBC membrane stabilisation of ethanol and ethyl acetate extracts are found to be lower when compared with the reference standard used, aspirin (94.27 ± 2.1 µg/ml).

Fig. 5.

Anti‐inflammatory activity based on HRBC membrane‐stabilisation assay. Values are means±SD, P˂0.05

3.3 Molecular docking of AKT1 (a serine/threonine–protein kinase) with bioactive compounds of ethanolic extract of CA

The docking study was accomplished with lung cancer cell signalling molecule AKT1 (also known as PKB) with active components of CA ethanolic extract (Fig. 2). Generally, protein kinases are involved in the regulation of signalling pathways, which contribute to the onset and progression of almost all types of cancers. The AKT pathway is one of the most important components of cell proliferation mechanism.

In the present paper, six out of eight ligands were selected from GC–MS analysis of CA ethanolic extract (Table 1) based on the percentage abundance (since the other two compounds are found in lower quantities) and docked with the X‐ray crystallographic structure of AKT1 (4EJN‐A) protein. A computational docking algorithm was used to predict the relative binding affinities for yohimbine, undecanoic acid 10‐methyl–ethyl ester, chrysin, oleic acid, oleic acidmethyl ester and oxiraneundeanoic 3‐phenyl–methyl ester trans‐ with AKT1, which allowed us to observe the structure‐inhibitory action relationships. The binding and interactions of all the above six compounds with AKT1 have been studied using molecular docking calculations. The binding of active components of CA ethanolic extract with AKT1 active site exhibited a very clear preference for the binding pocket. To prevent and treat several types of chronic cancers, suitable AKT1 inhibitors are to be investigated which is possible due to the identification and elucidation of AKT1 in the regulation of various metabolic pathways [29, 30]. In the current paper, the compounds yohimbine, undecanoic acid 10‐methyl–ethyl ester and chrysin have shown to have least binding energies of −9.94, −5.34 and −8.54 kcal/mol with two, one and four hydrogen bonds, respectively, with AKT1 (Table 2 and Figs. 6 a –c). The binding affinity of these compounds with AKT1 inhibitor is also high as evidenced from the docking score. Hence, the presence of the above active compounds might contribute to the anticancer property of the plant, whereas the interaction energy of other three compounds such as oleic acid, oleic acid methyl ester and oxiraneundeanoic 3‐phenyl–methyl ester trans‐ ranges from −4.85 to −5.92 kcal/mol with three and two hydrogen bonds, respectively (Table 2 and Figs. 6 d –f). The hydrogen bonds formed between the amino acids by appropriate distance for all the six active compounds have been shown in Table 2. Beevers and Huang [31] found that curcumin, a natural polyphenolic product of Curcuma longa, accomplish its anticancer property by primarily targeting AKT signalling and at high concentrations (>40 μM) it inhibits phosphorylation of AKT1 in many cancer cell lines. As evident from many reports, sulphoraphane administration prevents the progression of prostate cancer and pulmonary metastasis in transgenic adenocarcinoma of mouse prostate mice by minimising cell proliferation as a consequence of suppressing AKT signalling pathway [32].

Table 2.

Docking results of AKT1 with selective phytoligands of CA

Ligand name	Protein residue atom	Ligand atom	Distance, Å	Number of hydrogen bonds	Docking energy, kcal/mol
yohimbine	THR211‐O	H	2.27	2	−9.94
	THR211‐N	O	2.61
undecanoic acid,10‐methyl–methyl ester	THR211‐N	O	2.81	1	−5.34
chrysin	SER205‐OG	O	2.78	4	−8.54
	SER205‐OG	H	2.12
	SER205‐N	O	2.89
	SER205‐N	O	3.03
oleic acid	ARG328‐NH1	O	3.59	3	−4.85
	ARG328‐NH2	O	3.12
	ALA50‐N	O	2.78
oleic acid, methyl ester	GLN79‐NE2	O	3.01	2	−5.92
	THR82‐OG1	O	2.94
oxiraneundeanoic, 3‐phenyl–methyl ester, trans‐	SER205‐N	O	3.15	2	−5.54
	GLN79‐NE2	O	3.05

Fig. 6.

Interaction of ligands with AKT1. Docking conformation of

(a) Yohimbine, (b) Undecanoic acid 10‐methyl–ethyl ester, (c) Chrysin, (d) Oleic acid, (e) Oleic acid methyl ester, (f) Oxiraneundeanoic 3‐phenyl–methyl ester trans‐ with AKT1 receptor using AutoDock tool. The right image indicates the hydrogen bond interaction between AKT1 target and (a) Yohimbine, (b) Undecanoic acid 10‐methyl–ethyl ester, (c) Chrysin, (d) Oleic acid, (e) Oleic acid methyl ester and (f) Oxiraneundeanoic 3‐phenyl–methyl ester trans‐ using Accelrys Discovery Studio Visualiser

Allam et al. [33] demonstrated that computer‐aided drug design had been exploited to discover novel inhibitors through virtual screening and molecular docking of AKT1 and LMTK3 proteins by Dock Blaster server. To design and develop a novel inhibitor of AKT1, the author has also conducted a quantitative structure–activity relationship based on these compounds. A structure‐based virtual screening was performed using DOCK 4.0 programme and the X‐ray crystal structure of human Akt kinase to search for new inhibitors of Akt kinase. Out of 48 compounds tested, 26 compounds were showed more potent inhibitory effects on Akt kinase. These compounds were further evaluated for their cytotoxicity against HCT‐116 human colon cancer cells [34].

The phosphorylation of residues THR211 and SER205 is required for the activation of enzyme AKT1 [35]. Interaction of any ligand molecule with THR211 and SER205 prevents binding of the substrate to the active site, which can act as an inhibitor of AKT1. Accordingly, our present paper, reported that four out of six ligand molecules such as yohimbine, undecanoic acid, 10‐methyl–methyl ester, chrysin and oxiraneundeanoic and 3‐phenyl–methyl ester, trans‐ have been interacted with THR211 and SER205 (Table 2). Hence, these compounds possibly will perform the role as an inhibitor to AKT1 and conferring anticancer property to the plant CA. Mittal et al. [10] evaluated that the breast cancer cell signalling molecule Akt forms the binding site for cannabidiol, a natural compound with potential anticancer drug consists of GLU191, HIS194 and ASP274 residues at the binding site. The study compiled by Kumar et al. [36] demonstrated that the PDB‐derived AKT1 protein (PDB ID 4EJN) was subjected to molecular docking analysis using AutoDock software with 14 different native inhibitors of AKT1 derived from Selleck Chemicals website. The Akt1 exhibited a best binding affinity for two of the native ligands, namely Native_Akti‐1/2 (−13.2 kcal/mol) and Mutant Akti‐1/2 (−13.3 kcal/mol). The results of the present paper were comparable with the docking energy obtained for phytochemicals of CA such as yohimbine (−9.94 kcal/mol) and chrysin (−8.54 kcal/mol).

The factors analysed in the present paper are hydrogen bond interactions, binding energy, active site residues and orientations of ligands with receptors. These are measured as important keys of molecular docking, and if a compound exhibits lesser binding energy, the activity of the compound is high. Consequently, in the present paper, the docking scores on AKT1 indicated that there is a direct relationship between the energies of the binding affinity, referring to the lowest docking scores and stability. Among the six compounds, yohimbine exhibited the least binding energy (−9.94 kcal/mol) with two hydrogen bond interactions with AKT1 followed by chrysin (−8.54 kcal/mol) with four hydrogen bonds.

4 Conclusions

The overall picture of the present paper indicated that the ethanol extract exhibited the high potential of anti‐inflammatory activity. This paper corroborates the therapeutic prospective of CA suggesting that the plant could ‘lead’ for the isolation of new compounds with good efficiency to treat inflammation‐related disorders. Since the overexpression of Akt pathway is responsible for the development of lung cancer, it is necessary to initiate herbal medicines as a combination treatment targeting the inhibitors of Akt pathway. Molecular docking is one of such approaches and becoming popular nowadays to check the effectiveness of various drugs before beginning the clinical trials. The GC–MS‐derived components of ethanolic extract such as yohimbine followed by chrysin exhibited high binding affinity to AKT1. Molecular dynamics simulation studies can be performed further to examine the suitability of these compounds in cancer chemotherapy.