Exploring the therapeutic potential of Asparagus africanus in polycystic ovarian syndrome: a computational analysis

3.1.1 Primary sequence

The initial step involved the acquisition of the primary sequences of the target proteins, LHR and FSHR procured in FASTA format. This retrieval was conducted through UniProt database at (http://www.uniprot.org), utilizing the UniProt accession numbers P22888 and P23945 for LHR and FSHR, respectively. Notably, the primary sequence data indicated that LHR is comprised of 699 residues, while FSHR consists of 695 residues. The lengths of these proteins bear significance due to their potential influence on various functional aspects.

3.1.2 Physicochemical properties of proteins

To assess the physiochemical characteristics of LHR and FSHR, the ProtParam online tool (https://web.expasy.org/protparam/) was employed.

The findings of the physiochemical analysis (Table 1), revealed notable characteristics of the target proteins LHR and FSHR. The molecular weights of these proteins are determined to be 78,642.97 and 78,265.01, respectively. Moving to the isoelectric point (pI), LHR demonstrates a value of 8.82, indicative of its basic nature, while FSHR possesses a pI of 6.77, suggesting an acidic nature. A comparative analysis of charged residues reveals that LHR has a slightly lower count of negatively charged residues compared to FSHR. Conversely, in the realm of positively charged amino acids, LHR exhibits a higher quantity than FSHR. The extinction coefficients, assessed at a wavelength of 280 nm in water, provide insights into the proteins’ absorbance properties. For LHR, extinction coefficient 1 yields a value of 67,740 M⁻¹ cm⁻¹, while extinction coefficient 2 is calculated at 66,240 M⁻¹ cm⁻¹. In the case of FSHR, the corresponding values are 68,645 M⁻¹ cm⁻¹ and 67,270 M⁻¹ cm⁻¹, respectively.

Moving on to stability indices, LHR and FSHR present values of 39.74 and 43.41, respectively. Additionally, the aliphatic index, a measure of protein thermostability, is determined to be 100.06 for LHR and 108.73 for FSHR. Lastly, the GRAVY score, indicative of hydrophobicity, is calculated to be 0.171 for LHR and 0.227 for FSHR. These comprehensive physiochemical insights contribute significant understanding of the structural and functional properties of LHR and FSHR, offering valuable information for further investigations.

3.1.3 Protein 3D structure retrieval

In this study, the target proteins, LHR and FSHR, were sourced from the PDB (https://www.rcsb.org/) with entry IDs 7FII and 8I2H (Figure 1). The choice of PDB as the data source ensures the reliability and accuracy of the structural information. Notably, the advantage of utilizing these PDB entries is that the retrieved proteins, LHR and FSHR, do not necessitate additional purification steps. They are inherently devoid of undesirable molecules such as water molecules and ligands.

Figure 1:

3D structure of proteins. (A) Protein structure representing LHR and (B) protein structure representing FSHR.

3.2.1 Ligand chemical structure

A review of the literature led to the identification of 9 phytoconstituents associated with A. africanus [17]. The extraction of ligand 3D structures from A. africanus is executed through the utilization of two prominent databases, PubChem (https://pubchem.ncbi.nlm.nih.gov/) and ChemSpider (http://www.chemspider.com/) (Table 2).

Table 2:

Ligand structures.

Database	Ligand name/ID	Molecular formula	Molecular weight	Ligand structure
PubChem	FDHD/439368	C6H10O3	130.14 g/mol
	TMP/14296	C8H12N2	136.19 g/mol
	Phthalic acid/1017	C16H22O4	278.34 g/mol
	DTBO/545303	C17H24O3	276.4 g/mol
	Palmitic acid/985	C16H32O2	256.42 g/mol
	Stigmasterol/5280794	C29H48O	412.7 g/mol
	Sarsasapogenin/92095	C27H44O3	416.6 g/mol
ChemSpider	Butyl citrate/6261	C18H32O7	360.443 Da
	DMDP/510157	C28H46O2	414.664 Da

3.2.2 ADME properties

The SwissADME is an internet-based tool that computes significant pharmacokinetic, physicochemical, drug-like, and related parameters for one or many compounds. This is a valuable bioinformatics expertise for the initial drug development stages [20]. SwissADME (http://www.swissadme.ch/) is employed to evaluate ligands in order to identify the most suitable molecule for drug development.

The bioavailability radar gives an early assessment of the drug-likeness of the compounds of interest (Figure 2). The pink region serves as an illustration of the optimal physicochemical space for all predicted orally bioavailable properties. Polarity (POLAR), Lipophilicity (LIPO), Insolubility (INSOLU), Flexibility (FLEX) and Insaturation (INSATU) are the six physicochemical properties that are considered [21].

Figure 2:

Bioavailability radar of ligands.

Physicochemical characteristics (Table 3) include the molecular weight, molecular formula, number of heavy atoms, fraction csp3, number of aromatic heavy atoms, number of rotatable bonds, number of H-bond donors, number of H-bond acceptors, molar refractivity, and Topological Polar Surface Area (TPSA) [21].

Lipophilicity is an important element in drug discovery and design since it is the most informative and effective physicochemical parameter in medicinal chemistry (Table 4). SwissADME offers five publicly accessible models, namely WLOGP, XLOGP3, SILICOS-IT, MLOGP and iLOGP, to assess the lipophilicity of a compound. XLOGP3 is an atomistic approach that incorporates corrective factors and a knowledge-based library. WLOGP is an atomistic method that focuses on fragmental systems. MLOGP is a topological method that relies on a linear relationship with 13 implemented molecular descriptors. SILICOS-IT is a hybrid approach that uses 27 pieces and 7 topological descriptors. iLOGP is a physics-based approach that calculates the free energy of solvation in n-octanol and water using the generalized-born and solvent accessible surface area (GB/SA) model. The consensus log Po/w is calculated by taking the arithmetic average of the values predicted by the five proposed methodologies [21].

The water solubility (Table 5) prediction methods employed by SwissADME include the ESOL model [22], Ali [23], and SILICOS-IT. ESOL [22] is a model used to determine the solubility of a substance. The solubility is measured on a logarithmic scale, with values ranging from insoluble (<−10) to poorly soluble (<−6), moderately soluble (<−4), soluble (<−2), and highly soluble (0 or higher). Ali [23] solubility class is Log S, which indicates the scale of solubility. The scale ranges from insoluble (<−10) to poorly soluble (<−6), moderately soluble (<−4), soluble (<2), and highly soluble (<0). SILICOS-IT is classified based on solubility using the Log S Scale. It is categorised as insoluble when the Log S value is below −10, weakly soluble when it is between −10 and −6, moderately soluble when it is between −6 and −4, soluble when it is between −4 and −2, and highly soluble when it is between −2 and 0. SwissADME also offers measurements of solubility in mol L⁻¹ and mg mL⁻¹, as well as qualitative classes of solubility [21].

The BOILED-Egg (Figure 3) model is a spontaneous and efficient, approach to predict the passive gastrointestinal absorption, which is beneficial for drug discovery and development. The white region represents the area where molecules are more likely to be absorbed by the gastrointestinal tract due to their greater extent of absorption. On the other hand, the yellow region (yolk) indicates the space with the highest possibility of permeating to the brain. Cytochrome p450 (CYP) isoenzymes metabolise a significant portion (50–90 %) of medicinal compounds, utilising its five primary isoforms (CYP3A4, CYP2C9, CYP2D6, CYP2C19). P-gp is widely distributed throughout the intestinal epithelium, where it actively transports xenobiotics from the intestinal lumen back into the bloodstream. Similarly, P-gp is also present in the capillary endothelial cells of the brain, where it pumps xenobiotics from the brain back into the capillaries (Table 6) [21].

Figure 3:

BOILED-Egg representation. It shows passive gastrointestinal absorption (HIA) and brain penetration (BBB) with molecules in the WLOGP-versus-TPSA. Molecule 1; FDHD, molecule 2;TMP, molecule 3; phthalic acid, molecule 4; DTBO, molecule 5; palmitic acid, molecule 6; stigmasterol, molecule 7; sarsasapogenin, molecule 8; butyl citrate, molecule 9; DMDP.

Drug likeness evaluates the probability of a molecule to be administered orally as a drug, taking into account its bioavailability. Swiss ADME applies a process of screening chemical libraries to eliminate molecules that have characteristics that are not compatible with a desirable pharmacokinetics profile (Table 7). This is done using five distinct rule-based filters developed by major pharmaceutical companies, with the aim of enhancing the quality of their own chemical collections.The Lipinski filter, developed by Pfizer, is a pioneering set of criteria known as the rule of five [24]. The Ghose approach developed by Amgen, the Veber method developed by GSK, the Egan method developed by Pharmacia, and the Muegge method developed by Bayer [20].

After conducting an ADME study, a refinement process led to the exclusion of four out of the initial nine ligands due to demonstrated cytochrome (CYP) inhibition. Specifically, phthalic acid exhibited inhibitory activity on CYP1A2 and CYP2C19. Similarly, the compound DTBO demonstrated inhibitory effects on CYP2C19 and CYP2C9. Palmitic acid displayed inhibitory effects on both CYP1A2 and CYP2C9, while Stigmasterol exhibited inhibition specifically against CYP2C9.

Furthermore, DMDP was excluded from further consideration due to violations in four out of five druglikeness filters, namely Lipinski, Ghose, Egan, and Muegge. These filters are essential criteria in evaluating the drug-like properties of a compound.

With the refined ligand pool now consisting of four candidates – FDHD, TMP, sarsasapogenin, and butyl citrate – the next step involves evaluating their toxicity. This critical assessment aims to ensure that these remaining ligands not only exhibit desirable drug-like properties but are also devoid of potential harmful effects that could hinder their suitability for further development. The toxicity evaluation will contribute valuable information to guide subsequent decisions in the drug development process.

3.2.3 Toxicity prediction

The anticipation and assessment of potential harmful effects of compounds are integral components in the meticulous design and development of drugs. To address this critical aspect, the ProTox-II tool (https://tox-new.charite.de/), is employed for the purpose of conducting thorough toxicity analysis.

In the toxicity assessment of the four remaining ligands, it was revealed that three exhibited toxic effects. However, FDHD emerged as a non-toxic compound (Table 8). Consequently, given its favorable safety profile, the non-toxic ligand was deemed suitable for further in-depth analysis.

3.3.1 Drug structure retrieval

The freely accessible chemical database, ChemSpider (http://www.chemspider.com/), has been employed to retrieve the three-dimensional structure of Clomiphene citrate (Table 9).

Table 9:

Standard drug structure.

Standard drug	Molecular formula	Molecular weight	Drug structure
Clomiphene citrate	C32H36ClNO8	598.083 Da

3.3.2 ADMET properties

SwissADME (http://www.swissadme.ch/) is utilised for the analysis of the Absorption, Distribution, Metabolism, and Excretion (ADME) properties of Clomiphene citrate. SwissADME stands as a freely accessible tool that plays a pivotal role in providing comprehensive information across various domains crucial to medicinal chemistry. The tool offers insights into physicochemical factors (Table 10), lipophilicity (Table 11), water-solubility (Table 12), pharmacokinetics (Table 13), drug-likeness (Table 14), and other pertinent parameters [25]. By leveraging SwissADME, researchers gain valuable data that aids in the evaluation and optimization of potential drug candidates during the early stages of drug development.

Moreover, SwissADME goes beyond providing static information; it offers free access to a set of efficient predictive models, including the Bioavailability Radar (Figure 4), iLOGP, and BOILED-Egg (Figure 5) [20]. These predictive models enhance the tool’s utility by allowing researchers to anticipate crucial aspects such as bioavailability and lipophilicity, contributing to more informed decision-making in the drug development process.

Figure 4:

Bioavailability radar of clomiphene citrate.

Figure 5:

Boiled egg representation of clomiphene citrate.

In conjunction with SwissADME, the ProTox-II tool (https://tox-new.charite.de/) is employed to perform toxicity analysis specifically for the standard drug Clomiphene citrate, as outlined in Table 15. This dual-tool approach enables a comprehensive evaluation of both the physicochemical properties and potential toxic effects associated with Clomiphene citrate, providing a well-rounded understanding essential for advancing its development and application in medicinal contexts.

3.5.1 Target proteins and ligand

The selected target proteins are LHR and FSHR that are docked with FDHD.

LHR has the vina score of −5 and the ligand interacts with 6 amino acids of LHR target protein. Among these interactions, Methionine (MET) and Cysteine (CYS) contribute to Alkyl and Pi-Alkyl interactions, suggesting involvement in hydrophobic contacts and interactions with aromatic rings. Tryptophan (TRP), Tyrosine (TYR), and Arginine (ARG) are associated with Conventional Hydrogen Bonding, wherein hydrogen atoms are shared between the ligand and these amino acid side chains. TRP also represents Alkyl/Pi-Alkyl interaction. Serine (SER) exhibits a multifaceted interaction, participating in both Conventional Hydrogen Bonding and Carbon Hydrogen Bonding with the ligand. This dual interaction involves hydrogen bonds, with the additional presence of carbon atoms, enhancing the complexity of the ligand-protein binding (Figure 8).

Figure 8:

LHR target protein and ligand. (A) Molecular docking of LHR and ligand. (B) Protein interactions of LHR and ligand.

On the other hand, FSHR has the vina score of −4.4 and ligand interacts with 5 amino acids of FSHR. Leucine (LEU) is involved in Conventional Hydrogen Bonding, highlighting its role in forming hydrogen bonds with the ligand. Carbon Hydrogen Bonding is observed specifically with Alanine (ALA) at position A:486, indicating a unique interaction pattern involving carbon atoms. Furthermore, both ALA A:486 and A:463 contribute to Alkyl interaction, suggesting their participation in hydrophobic contacts. Additionally, Methionine (MET) and Valine (VAL) exhibit Alkyl interaction, signifying their involvement in the ligand-protein complex through hydrophobic interactions (Figure 9).

Figure 9:

FSHR target protein and ligand. (A) Molecular docking of FSHR and ligand. (B) Protein interactions of FSHR and ligand.

3.5.2 Target proteins and drug interaction

The target proteins LHR and FSHR are docked with standard drug clomiphene citrate, used for the management of PCOS.

LHR has the vina score of −5.9 and the ligand interacts with 6 amino acids of LHR target protein Valine (VAL), Lysine (LYS), Serine (SER), Arginine (ARG), and Tryptophan (TRP) all exhibit Conventional Hydrogen Bonding in their interactions with the ligand, indicating the formation of hydrogen bonds between the ligand and these amino acid side chains. Additionally, Carbon Hydrogen Bonding is observed specifically with Tyrosine (TYP) (Figure 10).

Figure 10:

LHR target protein and clomiphene citrate. (A) Molecular docking of LHR and clomiphene citrate. (B) Protein interactions of LHR and clomiphene citrate.

In contrast, FSHR has the vina score of −4.7 and ligand interacts with 4 amino acids of FSHR. Alanine (ALA) participates in Conventional Hydrogen Bonding, forming hydrogen bonds with the ligand. On the other hand, Isoleucine (ILE), Methionine (MET), and Leucine (LEU) are involved in Carbon Hydrogen Bonding. This interaction type entails hydrogen bonding with the ligand, with the additional presence of carbon atoms, contributing to the overall stability and specificity of the ligand-protein complex (Figure 11).

Figure 11:

FSHR target protein and clomiphene citrate. (A) Molecular docking of FSHR and clomiphene citrate. (B) Protein interactions of FSHR and clomiphene citrate.