IN-SILICO SCREENING OF PHYTOCHEMICAL COMPOUNDS IN CAESALPINIA BONDUCELLA L. SEEDS AGAINST THE GENE TARGETS OF OVARIAN STEROIDOGENESIS PATHWAY

ADME/T helps in identifying the potential drug candidates from a pool of drugs. Five phytochemical compounds, 33, 35, 38, 40, and 43 were found to have the ability to bind and inhibit appropriate targets in the ovarian steroidogenesis pathway. Hence, these compounds can be further characterized in vitro and in vivo for alleviating PCOS.


INTRODUCTION
Polycystic ovary syndrome (PCOS) is one of the most common endocrine disorders among women of reproductive age, and about 1 in 10 women are affected by this disorder worldwide (Meier 2018).PCOS is diagnosed using Rotterdam criteria wherein, a woman may present any two of the following three clinical symptoms: hyperandrogenism, oligo/anovulation and an ultrasound imaging of polycystic ovaries.In the long run, PCOS patients gradually acquire other metabolic risks such as hypertension, cardiovascular disease, type 2 diabetes and obesity (Boyle and Teede, 2016; Dokras et al., 2017).Based on their biochemical presentations, women with PCOS are categorized into two groups: hormonal imbalance (HI) phenotype with dysregulated hormonal cycle and metabolic syndrome (MetS) phenotype with abnormal metabolic values (Janani and Usha 2020).The MetS group, observed to be the significant phenotype among the Indian population, exhibits high levels of free testosterone indicating hyperandrogenism as the underlying pathology of PCOS.This group has high levels of blood glucose, insulin, cholesterol, HbA1C, and lower levels of High-density lipoprotein cholesterol (HDL-C).Hyper-insulinemic condition signals increased testosterone biosynthesis causing acne, obesity, Acanthosis nigricans and reduced follicle stimulating hormone (FSH) secretion and thus multiple premature follicles/cysts (Lerchbaum et al., 2014).In the hormonal imbalance phenotype, there is high estrogen (E2) level which raises luteinizing hormone (LH) secretion and decreases FSH secretion from the anterior pituitary.Low FSH level impedes the follicle stimulation, maturation and ovulation.Secretion of FSH can be enhanced by reducing the E2 levels with inhibitors of aromatase enzyme (CYP19A1 gene), that converts testosterone to estrogen (https://www.ncbi.nlm.nih.gov/gene/1588).Letrozole is the most widely used aromatase inhibitor with a 20-27% success in pregnancy (Holzer et al., 2006).FSH secretion from the pituitary can also be increased by blocking the estrogen receptors (ERα and ERβ) with selective estrogen receptor modulators (SERM) like clomiphene citrate, raloxifene and tamoxifen (Xu et al., 2021; Homburg 2005).In case of hyperandrogenism, treatment to reduce hirsutism, acne and infertility is targeted at lowering the activity of the enzyme 17 beta-hydroxysteroid dehydrogenases (encoded by HSD17B1 gene) that catalyzes the conversion of androstenedione to testosterone and/or by inhibiting the androgen receptors (https://www.ncbi.nlm.nih.gov/gene/3292).Flutamide and spironolactone are well-known antagonists of androgen receptors, while decanoic acid inhibits the HSD enzyme (Paris and Bertoldo, 2019).Ovarian steroidogenesis pathway and the effective inhibitors of androgen and estrogen biosynthesis is illustrated in the figure 1.Although widely used, these antiestrogenic and anti-androgenic drugs stance some side effects such as hepatotoxicity (Conway et al., 2014), hot flushes, nausea, vomiting, impotence, vision problem and osteoporosis in PCOS patients (Sidra et al., 2019).Hence, phytomedicines from natural sources are continuously being investigated for PCOS.
Figure 1 Representation of the inhibitory effect of commercially available drugs (red colour) on the genes and receptors of steroidogenesis pathway (blue colour).
Caesalpinia bonducella, a highly versatile medicinal shrub, is recently in the limelight to treat PCOS in India.C. bonducella L., also known as C. bonducella Flem.and C. crista Linn, belongs to the family of Caesalpiniaceae.It Polycystic ovary syndrome (PCOS) is the most common gynaecological disorder among reproductive-age women.Impaired metabolism of androgens and estrogens is one of the leading causes of PCOS.In India, medicinal herbs are being explored for their anti-androgenic and anti-estrogenic properties.In this study, we have screened the seed extracts of the herbal plant, Caesalpinia bonducella for potent inhibitors of estrogen and testosterone biosynthesis and assimilation.Methanol extract of C. bonducella seed kernels were subjected to gas chromatography -mass spectrometry (GC-MS) to identify the phytochemical constituents.Out of forty-three phytochemical compounds identified from the extract, eight compounds were selected based on Lipinski's rule of five for molecular docking.The selected phytochemical compounds were docked against specific targets of ovarian steroidogenesis pathway; human aromatase (CYP19A1), human 17β-hydroxysteroid dehydrogenase type 1 (HSD17B1), human androgen receptor and estrogen receptor α.Further, the nature of these compounds was validated using density functional theory (DFT) calculations and ADME/T studies.As per the molecular docking output, compounds 33, 35, 38, 40, and 43 exhibited higher binding affinities against the four selected targets.Phytochemical compounds were optimized using Gaussian 16 with the B3LYP function and the 6-31G(d, p) basis set and were correlated with docking results.ADME/T helps in identifying the potential drug candidates from a pool of drugs.Five phytochemical compounds,33,35,38,40,and 43 were found to have the ability to bind and inhibit appropriate targets in the ovarian steroidogenesis pathway.Hence, these compounds can be further characterized in vitro and in vivo for alleviating PCOS.
is commonly called bonduc nut, fever nut, and nicker nut (Khare 2007).The plant possesses anti-diabetic, hypolipidemic, antioxidant, anti-cancer, antipyretic and anti-inflammatory activities (Sachan et al., 2010;Sarma andDas, 2018, Shukla et al., 2009;Ihegwam et al., 2019;Shukla et al., 2010).More importantly, recent reports have demonstrated the antiestrogenic (Salunke et al., 2011) and antiandrogenic (Meerwal and Jain, 2016) potential of C. bonducella that has instigated us to identify the phytochemical compounds responsible for these activities.Molecular docking and the conceptual density functional theory (DFT) approaches were used for validating the docking results of the compounds identified by GC-MS from C. bonducella.The phytochemical compounds derived from C. bonducella were docked against the targets, human aromatase (CYP19A1), human 17β-hydroxysteroid dehydrogenase type 1 (HSD17B1), human androgen receptor and estrogen receptor alpha of ovarian steroidogenesis pathway.The interaction between target proteins and the phytochemical compounds were assessed based on their binding affinity.Further, the conceptual DFT and ADME/T studies were performed to render more knowledge on the chemical nature of these phytochemical compounds to facilitate further research on in vivo models.

Plant sample collection
C. bonducella seeds were collected from Kolli hills (GPS code -11°30'53.8"N78°14'24.9"E),Namakkal District, Tamil Nadu, India.The seeds were brought to the laboratory and stored at room temperature for further use.

Sample preparation and solvent extraction
The seeds collected were cleaned and the outer seed coat was removed.The inner seed kernels that are widely used in the treatment of PCOS were shade dried for 10 -15 days.Fifty grams of the ground seed kernels were added to 250 ml of methanol and was extracted using soxhlet apparatus for ten hours (5 cycles/hour) at 60 °C.The crude extracts obtained were filtered and the resulting solution was concentrated in a rotatory vacuum evaporator at 55 °C.The resultant crude sample was stored at 4 °C for GC-MS analysis and further experiments.

GC-MS analysis
Phytochemical investigation of seed kernel extracts of C. bonducella was performed in GC 7890B equipment connected to 5977A MS equipped with an autosampler (Agilent Technologies, USA).Concentrated and filtered extract of concentration 1 mg/ml, dissolved in methanol, was used for GC-MS analysis.The experimental conditions of GC-MS were as follows: HP-5MS 5% phenyl methyl siloxane capillary column with a dimension of 30 μm × 250 μm × 0.25 μm.The flow rate of the mobile phase with carrier gas helium was set at 1 ml/min with 70ev ionization energy in an electron impact mode.The sample injection volume was 1 μl, and the temperature programme (oven temperature) was 60 °C raised to 280 °C at the rate of 5 °C/min.The database namely the National Institute of Standards and Technology (NIST2011) was used to identify the unknown compounds by comparing them with the standard compounds reported earlier.

Preparation of protein
The 3D structures of the target proteins human aromatase (CYP19A1) (PDB ID: 3S79), human 17β-hydroxysteroid dehydrogenase type 1 (HSD17B1) (PDB ID: 1FDS), human androgen receptor (PDB ID: 1E3G) and estrogen receptor α (PDB ID: 3ERT) were retrieved from the protein databank (PDB) (https://www.rcsb.org/).The 3D structure of the target proteins was visualized in PyMol software (http://www.pymol.org)and the H2O molecules and associated ligands were removed initially.Furthermore, the target proteins' energy was minimalized by adding charges to it using Autodock vina in PyRx software and the protein structure obtained was converted to pdbqt format for docking

Ligands selection
The drug likeliness property of all the phytochemical compounds was investigated by the best-known physical property filters Lipinski's rule of five, using (https://www.scfbio-iitd.res.in/utility/LipinskiFilters.jsp)webserver.Lipinski's rule of five (molecular mass, number of hydrogen bond donors, and acceptors, log P and molar refractivity) was subjected to ligand preparation.

Preparation of ligands
The compounds selected based on Lipinski's rule of five were retrieved from the PubChem chemical database (https://pubchem.ncbi.nlm.nih.gov/).The ligand structures were uploaded to the PyRx open source software with OpenBabel plugin.The Spatial data file (SDF) of the ligands were converted to PDB format using the OpenBabel software (O'Boyle et al., 2011).Further, the structure of the ligands was prepared by assigning chargers, identifying the torsion root, correcting its angles and optimizing it using UFF (Universal force field).The generated ligand structures were converted to pdbqt format in order to generate 3D atomic coordinates of the molecules.

Active site prediction
The ligand can bind to many sites, but suitable ligand binding sites are defined from the ligands attached to the original target protein's active site.Active site residues were defined as atoms situated less than 25.0 °A from any ligand atom (Ashraf et al., 2014).To identify the active sites, the structures of the target proteins were uploaded to the active site prediction server maintained by Supercomputing Facility for Bioinformatics and Computational Biology, Indian Institute of Technology, Delhi, to predict the probable ligand binding sites (http://www.scfbio-iitd.res.in/dock/ActiveSite.jsp).

Molecular docking
In this study, molecular docking was performed using Autodock vina in PyRx virtual screening open-source software maintained by the National Biomedical Computation Resources (Stouten and Kroemer;2006;Kitchen et al., 2004;Vieira and Sousa, 2019).Four proteins and eight phytochemical compounds were selected under the vina wizard control for molecular docking.A grid appears over the target protein structure, and the grid size is adjusted based on the active site residues chosen.The docking was performed by Autodock vina and the strength of the interaction of the compounds and the target proteins was predicted based on the binding energy scores.

Conceptual DFT
Density functional theory (DFT) is a computational quantum mechanical modelling tool for examining the electronic structure of receptors and ligands, as well as the interactions between them.The electronic and structural parameters of the top ten hits were estimated using the B3LYP method and the 6-31G (d, p) basis set, with Gaussian 09 (Jordaan et al., 2020).Lowest unoccupied molecular orbital (LUMO), total energy, highest occupied molecular orbital (HOMO), global softness, energy gap, absolute hardness, molecular dipole moment, electrophilicity index, electronegativity and chemical potential were among the estimated parameters evaluated (Kulkarni et al., 2020).

ADME/T analysis
Poor ADME/T and pharmacokinetic properties can lead to the drug failure in experimental studies in the beginning or at lateral stage.So,early prediction of ADME/T properties can be useful in drug discovery (Panwar and Singh, 2021).The SDF (Structure Data Format) files and SMILES (Simplified Molecular-Input Line-Entry System) of the phytochemical compounds were retrieved from PubChem database for ADME/T analysis.The ADME/T (absorption, distribution, metabolism, excretion, and toxicity) properties of the chosen ligands were detected to understand their pharmacokinetic properties and their pharmacodynamic effects on the body.Analysis of ADME/T properties was done using the SwissADME (http://www.swissadme.ch/) and ADMET lab server (http://admet.scbdd.com).

GC-MS analysis
After successful conventional soxhlet extraction, GC-MS analysis led to the identification of various compounds in the seed kernels of C. bonducella.The GC-MS chromatogram (Figure 2) recorded a total of 49 peaks which indicated the presence of 43 bioactive compounds with different retention times and percentage peak areas (Table 1).
Figure 2 The GC-MS chromatogram of the methanol seed kernel extract of C. bonducella.The peak area percentage that depicts the weight percentage ratio of the particular compound in the mixture was investigated, and eight phytochemical compounds were found to be abundant.Among the identified compounds, compound 21 had the maximum area percentage of 28.

Selection of the ligands
From the GC-MS analysis, only eight out of forty-three phytochemical compounds satisfied the Lipinski rule of five.The main goal to verify Lipinski rule is to obtain a preliminary prediction of the potential pharmacological capacities of a molecule to become a drug (Bouzekri et al., 2022).Therefore, the compounds listed in Table 2 were used as ligands for molecular docking with target proteins.Investigation of drug-likeness properties accelerates the process of drug discovery and development.

Molecular docking
The selected compounds were docked with the target proteins human aromatase (CYP19A1), human 17β-hydroxysteroid dehydrogenase type 1 (HSD17b1), human androgen and estrogen receptor α) to identify the binding affinities of the compounds.In docking, the RMSD value is used to compare the docked conformation to the reference or other docked conformations.During the docking analysis, the most least RMSD value was selected from all conformations generated.The ligand and protein complex with hydrogen and hydrophobic interactions and binding energies are shown in Table 3. Compound 43 -5.0 His524 Leu346, Leu349, Trp383, Leu384, Leu391, Phe404, Leu525 The compounds selected for visualization based on docking results are mentioned in bold Protein docking results of human aromatase (CYP19A1) revealed compound 35 (-8.7 kcal/mol), compound 39 (-8.8 kcal/mol) and compound 43 (-9.0 kcal/mol) to have a good binding affinity (with hydrogen bond and hydrophobic bond interaction) with a better docking score among the eight compounds.Compounds 35, 39, 43 showed hydrogen bonding interaction with the amino acid residues Arg115, Thr310, Thr310, Ser314, Val370, Ala438 and Gly439.The docked position of individual top three ligands in human aromatase (CYP19A1) is shown in figure 7.For better visualisation, they have been represented in different colours.Compounds 35, 39 and 43 showed potential binding affinities to the targets.Additionally, these compounds also had better binding scores and hydrogen bond interactions.Another study also demonstrated that the native ligand of androstenedione (ASD) docked against the human aromatase (CYP19A1) protein in polar and non-polar interactions, matched our present findings (Kumavath et al., 2016).Figure 8 (a) The binding poses of compound 33 (red), compound 35 (blue) and compound40 (brown) to human 17-beta-hydroxysteroid-dehydrogenase type 1 (HSD17b1) (b) compound 33 interacting with the residues, Ser142 and Tyr155 (green) forming hydrogen bonds (yellow) of length 2.3 Å and 2.0 Å, respectively (c) compound 35 interacting with the residues, Asn152 and His221 (green) forming hydrogen bonds (yellow) of length 2.5 Å and 1.8 Å, respectively (d) compound 40 interacting with the residues, Tyr218 (green) forming hydrogen bonds (yellow) of length 2.0 Å.Further, docking of human androgen receptor revealed that the compound 33 (-8.3 kcal/mol), compound 35 (-7.0 kcal/mol) and compound 38 (-5.6 kcal/mol) showed good binding affinities with better docking scores among all the ligands.Amidst the produced conformations, with the amino acid residues Asn705, Gln711, Arg752, and Thr877.The binding poses of the top ligands in the active site of the human androgen receptor are depicted in figure 9. Other phytochemical compounds showed low binding affinity and the absence of hydrogen bonding.Compound 33 could be an effective anti-androgen because the hydrogen bond interacting amino acid residue (Thr877) observed in the active site of the human androgen receptor in this study was similar to the active site residues of the human androgen receptor followed on docking the piperine, which showed an antagonistic effect on dihydrotestosterone and proved to be an effective male contraceptive pill (Chinta et al., 2015).The ethanolic seed extract of bonduc was treated to male wistar rats for 60 days in dose dependent manner (200 and 400 mg/kg) to check the anti-androgenic activity and after 60 days, the sperms of rats got affected and a decrease in testosterone levels observed.There is a reduction in sperm count, motility and bulging of acrosomal membrane and detachment of sperm heads happens due to androgen deficiency.Thus, there is an action to reduce the hyperandrogenism which is a main factor for PCOS (Kandasamy and Balasundaram, 2021a).
Figure 9 (a)The binding poses of compound 33 (red), compound 35 (blue) and Compound 38 (teal) to human androgen receptor (b) compound 33 interacting with the residues, Thr877 (green) forming hydrogen bonds (yellow) of length 2.7 Å and 3.2 Å, respectively (c) compound 35 interacting with the residues, Asn705, Gln711, Arg752 (green) forming hydrogen bonds (yellow) of length 2.6 Å, 2.4 Å and 1.6 Å respectively (d) Compound 38 interacting with the residues, Gln711 and Arg752 (green) forming hydrogen bonds (yellow) of length 2.5 Å and 2.6 Å.On docking the ligands to the target estrogen receptor α, compound 33 (-8.6 kcal/mol), compound 35 (-8.0 kcal/mol), and compound 38 (-8.8 kcal/mol) represented good binding affinity with better docking score among the eight compounds.Arg394, His524 and Glu353 demonstrated hydrogen bonding.The binding poses of the top ligands in the active site of estrogen receptor α are depicted in figure 10.Compound 35 could be a better anti-estrogenic compound because the amino acid residues through which this compound interacts with the target estrogen receptor α is similar to the residues found in the active site of 4-hydroxytamoxifen -ERα interaction.4-hydroxytaxoxifen drug was prescribed for ER-positive breast cancer patients that showed an effective binding score to estrogen receptor alpha (Acharya et al., 2019).For checking the anti-estrogenic activity, the female albino rats were used and they were treated with bonduc seeds for 10 days.The results showed that there is a significant effect on follicles, reduction of weight of ovary and uterus and endometrial degeneration occurs.These results showed that there is an anti-estrogenic activity in the bonduc seeds (Kandasamy and Balasundaram, 2021b).

Conceptual DFT
Table 4 shows the statistics of DFT-based molecular descriptors of chosen phytochemical components.Following the optimization of the eight compounds with the B3LYP method and based on molecular orbital thery, ten different descriptors were generated (Nagarajan et al., 2020a).The energies of the LUMO (ELUMO) and HOMO (EHOMO) molecular orbitals were estimated using Fukui's theory.Here, we found that compounds (33, 35, 38, 39, 40, 41 and 43) can easily metabolise cytochrome P450 family enzyme.Selected eight phytochemical compounds shows that less than 3 hours it denotes the lower level of half-life and these compounds easily eliminate from our body.The criteria that high biological activity through lower toxicity of compounds elementary condition for drug discovery process.The phytochemical compounds contain a high level of oral acute toxicity and carcinogenicity to eliminate for future study (Rahman et al., 2021).Toxicity evaluation of selected phytochemical compounds prediction represents that except compound 42, all the other compounds (33, 35, 38, 39, 40, 41 and 43) are non-blockers of hERG and there is no cardiac side effect in humans.AMES (Ames Mutagenicity) shows that six compounds (33, 35, 38, 39, 40 and 43) respond to non-mutagenicity and did not cause any genetic damage.

CONCLUSION
PCOS is a heterogeneous disorder with life-threatening long-term complications.
Therefore, this health issue should be addressed immediately using alternate therapies in this study, seed kernels of C. bonducella was screened for potential anti-estrogenic and anti-androgenic phytochemical compounds by computational analysis.The results of this study suggest that the phytochemical compounds (33, 35, 38, 40 and 43) showed good binding affinities to the targets human aromatase (CYP19A1), human 17β-hydroxysteroid dehydrogenase type 1 (HSD17b1), human androgen and estrogen receptor α) of ovarian steroidogenesis pathway.Hence, these compounds could be recommended as anti-estrogenic and anti-androgenic drugs in the management of PCOS.On the whole, the current study provides valuable insights and opens the door to new research opportunities for the exploration of the phytochemical compounds in in vitro and in vivo PCOS models.
Human aromatase (PDB: 3S79), human 17β-hydroxysteroid dehydrogenase type 1 (PDB: 1FDS), human androgen receptor (PDB ID: 1E3G) and estrogen receptor α (PDB: 3ERT) active sites as recommended by the literature matched with the SCFBio active site prediction server as well.Our results show that the binding active site residues of human aromatase (CYP19A1) were shown in figure3.Human 17β-hydroxysteroid dehydrogenase type 1 (HSD17b1) active site residues are clearly shown in figure4, 5 and 6 shows the binding active site residues of human androgen and estrogen receptor α respectively.

Figure 3
Figure 3 Active site for ligands in human aromatase (CYP19A1).The bind site representing the (a) surface, (b) chains and (c) cavity enlarged to unveil the amino acid residues interacting in <5.0 Å vicinity.

Figure 4
Figure 4 Active site for ligands in human 17β-hydroxysteroid dehydrogenase type 1 (HSD17b1).The bind site representing the (a) surface, (b) chains and (c) cavity enlarged to unveil the amino acid residues interacting in <5.0 Å vicinity.

Figure 5
Figure 5 Active site for ligands in human androgen receptor.The bind site representing the (a) surface, (b) chains and (c) cavity enlarged to unveil the amino acid residues interacting in <5.0 Å vicinity.

Figure 6
Figure 6 Active site for ligands in estrogen receptor α.The bind site representing the (a) surface, (b) chains and (c) cavity enlarged to unveil the amino acid residues interacting in <5.0 Å vicinity.Molecular docking

Figure 11
Figure 11 DFT -Electron density maps of the selected eight phytochemical compounds from methanol seed kernel extract of C. bonducella ADME/T analysis Chemical absorption, distribution, metabolism, excretion, and toxicity (ADMET) are important factors in drug discovery and development.A high-quality drug candidate should not only be effective against the therapeutic target, but it should also exhibit appropriate ADMET properties at a therapeutic dose (Guan et al., 2018).ADME/T (absorption, distribution, metabolism, excretion, and toxicity) and

Table 1
GC-MS data of the phytochemicals present in C. bonducella methanol seed kernel extract

Table 2
Compounds selected based on Lipinski's rule of five

Table 3
Docking score of the selected eight phytochemicals

Table 4
DFT descriptors of the selected eight phytochemical compounds from C. bonducella methanol seed kernel extract Compound 41 had the lowest ΔE, while compound 42 had the highest ΔE.The narrower the energy gap, the greater the molecule's activity, which is linked to the transition of molecules from the HOMO to the LUMO state (

Yadalam et al., 2021b).
Low ΔE values were also found in compounds 40 and 35.The energy gaps (E) between the HOMO and LUMO of the molecules were calculated using the difference in energies between the two orbitals: ΔE = ELUMO -EHOMO (Nagarajan et al., 2020b).Figure11shows electron density maps that show the density of electrons in various locations of the molecules.A molecule's chemical reactivity is proportional to its molecular dipole moment (

Yadalam et al., 2021c; Mert et al., 2011).
Compound 40 has the maximum dipole moment (12.2081debye), followed by compound 35 (8.6926 debye) and compound 33 (6.8689debye).The ability of a molecule to take electrons is measured by its electronegativity.It is a crucial criterion for evaluating how effective the molecule's inhibition is.If a molecule's electronegativity is lower, it will have a greater effect on inhibiting other molecules (

Yadalam et al., 2021d).
The electronegativity index of Compound 40 (-6.01) is the lowest, followed by compound 35 (-5.655) and compound 41 (-5.5).The results of conceptual DFT are compatible with the docking results of the selected compounds.

Table 5
ADME/T properties of the selected eight phytochemical compounds from C. bonducella methanol seed kernel extract Properties

Glue and Clement, 1999).
Here, selected eight phytocompounds have shown good human intestinal absorption (HIA) and Caco-2 permeability values.ADME/T is a critical parameter for determining a compound's fate when taken orally.In vitro permeability studies are commonly done in the Caco-2 cell line, which resonates the compound's intestinal absorption (Natesh et al., 2021).Following absorption, compounds travel via the bloodstream to different regions of the body before being processed in the liver.Cytochrome P450 family containing a set of enzymes can break down ingested drug molecules in the liver and remove them through bile and urine (