Evaluation of natural compounds as folate biosynthesis inhibitors in Mycobacterium leprae using docking, ADMET analysis, and molecular dynamics simulation

1 Introduction

Leprosy is a persistent infectious disease caused by Mycobacterium leprae. It can affect the muscles, eyes, skin, and nose in addition to damaging peripheral nerves. In leprosy, nerve damage can result in severe incapacitating deformities [1]. Leprosy continues to be a major global health concern, especially in underdeveloped areas. Presently, the infection is present in more than 100 countries, frequently in high-burden regions with a low-burden case background [2]. The host’s immune reaction to the bacterium is reflected in a variety of clinical manifestations. Although M. leprae is susceptible to many contemporary antimicrobial agents, prolonged therapy is necessary for effective treatment [3].

Drug resistance to the main therapeutic agents appears to be low, though dapsone, rifampin, and ofloxacin have all been known to exhibit single-drug resistance. Novel approaches are required to develop antileprosy agents due to the challenges posed by the current therapeutic regimens and the emergence of drug-resistant strains [4,5].

An important target in the hunt for potent antileprosy drugs is the FolP protein. The FolP protein, also called dihydropteroate synthase (DHPS), is a key component of the folate biosynthesis pathway, which is a vital metabolic process that is necessary for M. leprae and other microorganisms to survive. Folate, also known as vitamin B9, is a precursor to several molecules found in cells, such as nucleic acids, amino acids, and cofactors in one-carbon transfer reactions [6].

The FolP protein plays a critical role in the synthesis of dihydropteroate, a precursor required for the formation of tetrahydrofolate (THF), in the case of M. leprae. THF is a coenzyme that helps transfer one carbon atom when nucleotides and amino acids are being biosynthesized. The bacterium’s capacity to synthesize nucleic acids and other crucial cellular components is hindered by the inhibition of FolP, which disrupts this synthetic pathway and depletes relevant folate derivatives [7].

FolP’s mechanism of action involves the condensation of p-aminobenzoic acid (PABA) with 6-hydroxymethyl-7,8-dihydropterin-pyrophosphate, resulting in dihydropteroate formation. This reaction is critical in the de novo synthesis of folate because it serves as a precursor for subsequent enzymatic reactions that result in the formation of fully active folate derivatives [8,9].

FolP has become an appealing target for the development of antimicrobial agents, including those aimed at treating leprosy, due to its critical role in the folate biosynthesis pathway. FolP inhibition disrupts the supply of essential folate cofactors, eventually stopping bacterial growth and making the pathogen vulnerable to the host’s immune defenses [10].

Searching for compounds that can specifically target and inhibit the activity of this enzyme often entails screening chemical libraries, including those containing natural compounds. Natural compounds have a number of benefits, such as their structural diversity, bioavailability, and, frequently, a long history of safe use [11,12]. This study’s objective is to look at the potential of natural compounds as FolP protein inhibitors in M. leprae. Computational approaches such as virtual screening, molecular docking, and dynamics simulations are used to predict the binding affinities and interactions between potential inhibitors and the FolP protein.

2 Methodology

3 Results

3.1 Structure validation

The quality of the 3D structure of the FolP protein predicted by AlphaFold was assessed using the ERRAT quality factor and Ramachandran plots. A reliable model typically exhibits an average overall quality factor of approximately 91% for the ERRAT score. The ERRAT quality factor for FolP was found to be 97.72% (Figure 1). Additionally, the Ramachandran plot for FolP indicated that all residues were located in the favored region, with no residues falling in the disallowed region (Figure 2). According to MolProbity results, the model had excellent side‐chain quality (0% poor rotamers and 99.04% favored rotamers, only 7 out of 2,882 (0.24%) vs a goal of <0.1% bad angles, and just 3 CaBLAM outliers (1.1%) which is just above the desired threshold (<1.0%). All of these problems in the 3D structure of FolP were solved during receptor preparation with the Protein Preparation Wizard which improved its backbone geometry more [18].

Figure 1

ERRAT quality factor of the predicted structure of the FolP protein.

Figure 2

Ramachandran plot of the FolP protein. The yellow regions show the allowed regions, while the white regions show the disallowed regions.

3.2 Molecular docking

The natural compounds were prepared, and a docking study was conducted against the FolP protein [24]. The binding affinities of all docked compounds were analyzed, and then top ten compounds with binding affinities ranging from −7.851 to −7.219 kcal/mol were selected for further analysis (Table 1). The docking scores of the selected compounds suggested that these have potential for inhibiting the function of the FolP protein.

Table 1

Glide scores of the docked compounds against the FolP protein

Sr.	Compounds	Structures	Glide score (kcal/mol)
1	LTS0073756		−7.851
2	LTS0262854		−7.848
3	LTS0076695		−7.62
4	LTS0235062		−7.601
5	LTS0035466		−7.594
6	LTS0174667		−7.456
7	LTS0241035		−7.429
8	LTS0211389		−7.421
9	LTS0172379		−7.294
10	LTS0033598		−7.219

The docked poses of the selected compounds were analyzed using the Discovery Studio client tool to find the molecular interactions. The molecular interactions mainly involved hydrogen bonding, van der Waals interactions, pi–pi stacking, pi–sigma interactions, and alkyl (hydrophobic) interactions. The molecular interactions of each candidate compound help in determining the binding affinities. Particularly, the hydrogen bonds between the ligand and protein atoms play an important role in the strength of protein–ligand complexes [25]. LTS0073756 formed seven conventional hydrogen bonds with Asn17, Tyr141, Gly181, Arg253, Arg54, Arg233, and His255 and three alkyl interactions with Ala57, Pro55, and Arg214. LTS0262854 formed nine conventional hydrogen bonds with Asn189, Phe215, Asp86, Thr53, Arg54, Arg212, Arg233, Ala183, and Tyr141; one pi–cation interaction with Arg253; and two alkyl interactions with Arg214 and Pro55. Similarly, LTS0076695 made four conventional hydrogen bonds with Asp86, Arg54, Arg214, and Arg212; one pi–cation interaction with Arg253; and three alkyl interactions with Lys213, Pro55, and Pro230. LTS0235062 formed eight conventional hydrogen bonds with Gly181, Arg212, Arg214, His255, Arg54, Asn13, Thr53, and Asp86; one pi–cation interaction with Arg253; and two alkyl interactions with Pro55 and Lys213. The molecular interactions of the remaining compounds are shown in Table 2.

Table 2

Molecular interactions of the selected compounds with the FolP protein

Sr.	Compound code	Interactions
1	LTS0073756	Conventional hydrogen bond: Asn17, Tyr141, Gly181, Arg253, Arg54, Arg233, and His255
		Alkyl: Ala57, Pro55, and Arg214
2	LTS0262854	Conventional hydrogen bond: Asn189, Phe215, Asp86, Thr53, Arg54, Arg212, Arg233, Ala183, and Tyr141
		Pi–cation: Arg253
		Alkyl: Arg214 and Pro55
3	LTS0076695	Conventional hydrogen bond: Asp86, Arg54, Arg214, and Arg212
		Pi–cation: Arg253
		Alkyl: Lys213, Pro55, and Pro230
4	LTS0235062	Conventional hydrogen bond: Gly181, Arg212, Arg214, His255, Arg54, Asn13, Thr53, and Asp86
		Pi–cation: Arg253
		Alkyl: Pro55 and Lys213
5	LTS0035466	Conventional hydrogen bond: Lys213, Ser211, Arg253, Thr53, Asn13, and Arg54
		Alkyl: Pro230, His255, Arg212, Phe182, and Pro55
6	LTS0174667	Conventional hydrogen bond: Thr53, Asn13, His255, Arg54, Tyr141, Arg212, and Gly181
		Pi–sigma: Ala57
		Alkyl: Lys213 and Pro55
7	LTS0241035	Conventional hydrogen bond: Arg253, Tyr141, Arg214, and Asn17
		Carbon–hydrogen bond: Gly181, Pro55, Arg54, and Ala57
		Alkyl: Lys213
8	LTS0211389	Conventional hydrogen bond: Arg214, Arg54, Arg253, Asp86, and Thr16
		Carbon–hydrogen bond: Tyr141, Arg212, His255, Thr53, Asn13, and Ser52
		Alkyl: Lys213 and Pro230
9	LTS0172379	Conventional hydrogen bond: His255, Arg54, Gly181, Arg214, Ile58, and Asn17
		Carbon–hydrogen bond: Lys213 and Pro55
		Alkyl: Ala57
10	LTS0033598	Conventional hydrogen bond: Arg253, Tyr141, Asp16, Arg54, and Arg214
		Carbon–hydrogen bond: Gly181, Ala57, Pro55, and Asn17
		Alkyl: Lys213

3.4 Binding pose analysis

The plausible binding modes of three compounds with better drug-likeness and drug scores were observed. This analysis showed that the docked poses of the hits occupied the same space in binding sites of the FolP protein (Figure 3). As a result, the plausible binding modes of the docked hits were evaluated for stability through MD simulations.

Figure 3

Plausible binding modes of three selected compounds along with molecular interactions with FolP binding pocket residues.

3.5 MD simulations

3.5.1 RMSD

The RMSD of the Cα atoms was measured to evaluate the structural changes of the complexes [29,30]. For the LTS0262854 complex, the RMSD values gradually increased to 3.5 Å at 80 ns, and they remained stable in this range until 120 ns. The RMSD increased to 4 Å after 120 ns and stayed in this range until the simulation end. At 20 ns, the RMSD of LTS0241035 increased to 4 Å, and it stayed in this range throughout the time. The RMSD of protein alpha atoms in the LTS0033598 combination was stable for 40 ns before progressively increasing to 5 Å at 130 ns. It again decreased to 4 Å at 160 ns and then maintained this range (Figure 4).

Figure 4

RMSD plots of the FolP complexes.

3.5.2 RMSF

RMSF values were calculated to investigate the protein residue dynamics upon interacting with the ligands [31]. The RMSF plots revealed that most of the residues did not show many fluctuations during simulation as the RMSF values were lower than 2 Å throughout the simulation, indicating that the ligands did not exert the fluctuations in the protein. In contrast, the residues in the loop regions showed high fluctuations in the presence of hit compounds, as compared to the co-crystal ligand complex, reaching around 7 Å, suggesting increased flexibility in these regions (Figure 5). According to RMSF analysis, the protein–ligand complex exhibited overall stability, as most residues maintained a rigid conformation.

Figure 5

Comparative RMSF plots of the FolP protein.

3.5.3 Protein–ligand contacts

The interactions between the atoms of protein and ligands were observed during the simulation, and it was observed that the main interactions involved hydrogen bonding, water bridges, ionic interactions, and hydrophobic interactions. These contacts play an important role in protein–ligand complex stability during the simulation. Residues that form hydrogen bonds with LTS0262854 were Asn13, Thr53, Asp86, Trp132, Met135, Ala137, Ala183, Arg112, Lys213, Arg214, Arg233, and Arg253. In the LTS0241035 complex, the hydrogen bonding interactions involved Asn13, Asn17, Ser20, Ser52, Arg54, Thr53, Asp86, Tyr141, Glu142, Ala143, Ala183, Arg214, and Arg218. Finally, the hydrogen bonding in LTS0033598 was observed in the following residues: Arg89, Gly109, Gly110, Arg111, Pro114, Leu129, Trp132, Leu134, Asp156, and Gln163 (Figure 6). The hydrogen bonding during MD simulations provides important insights into the specific residues that stabilize the protein–ligand complexes. This information enhances our understanding on how these interactions contribute to the complex stability and binding affinities [32].

Figure 6

Protein–ligand interactions. Green bars show hydrogen bonding, gray bars show hydrophobic interactions, and blue bars show water bridges.

3.5.4 Solvent accessible surface area (SASA)

An investigation was conducted to find the protein’s SASA during simulation and to search for any conformational changes. The analysis revealed that the protein had an initial SASA value around 12,500 Ų, and it remained in the range of 12,500–13,500 Ų in all complexes. In the LTS0033598 complex, the SASA values reached 14,000 Ų for some time, but it again attained the stable range of 12,500–13,500 Ų (Figure 7). The SASA values indicated that the protein structure did not face conformational changes during the simulation when these ligands were bound to it.

Figure 7

SASA plots of the FloP complexes with selected compounds.

The detailed chemical interactions particularly observed the formation of stable hydrogen bonds and hydrophobic contacts with key active site residues strongly suggest that the selected ligands can effectively interfere with the binding of the natural substrate, PABA. This interference is expected to inhibit the folate biosynthesis pathway by reducing the production of THF, thereby compromising nucleotide synthesis and bacterial proliferation. Clinically, such enzyme inhibition could translate into significant reduction in the bacterial load in patients with leprosy, potentially leading to more effective treatments with a lower likelihood for developing drug resistance.

4 Discussion

Natural compounds found in plants, microbes, and marine organisms represent an untapped reservoir for novel drug discovery. This research will involve screening a library of natural compounds chosen for their structural features and known bioactivity in order to identify potential FolP inhibitors. Computational modeling and bioinformatics analyses will supplement experimental assays by providing insights into the binding mechanisms and interactions between the selected compounds and the FolP protein [33,34].

FolP inhibitors may be useful tools in unraveling the complex biology of M. leprae in addition to their potential as anti-leprosy agents. Understanding the structural dynamics of the FolP protein and its interactions with natural compounds can help rationally design targeted therapeutic interventions.

The purpose of this study was to describe the virtual screening of natural products against the FolP protein of M. leprae to find the best hits for wet lab studies targeting M. leprae. The LOTUS database was used to obtain a natural product library containing 1,400 compounds. The ligand structures were prepared for further study using the LigPrep program. AlphaFold predicted the 3D structure of the FolP protein, which was then validated by calculating the ERRAT quality factor and analyzing the Ramachandran plot.

Using the standard precision mode of the Glide tool, the library of natural products comprising 1,400 compounds was docked against the FolP protein. Ten of the most promising compounds were chosen for additional examination based on their binding affinities. The chosen compounds’ binding affinities ranged from −7.851 to −7.219 kcal/mol. The chosen compounds’ binding affinities suggested that they may potentially inhibit the FolP protein’s function. Clinically validated inhibitors of DHPS (FolP), including sulfonamides such as sulfamethoxazole and dapsone, have historically exhibited moderate binding affinities in computational and experimental studies. According to a previous study, pterin-based inhibitors targeting the FolP enzyme have shown docking scores generally ranging from 4.91 to –8.69 kcal/mol including the control group compound inhibitors, with some of the most potent inhibitors displaying IC₅₀ values in the micromolar range (e.g., compound 2, IC₅₀ = 19.8 µM) against Bacillus anthracis DHPS. Despite their effectiveness, many of these inhibitors are hindered by solubility issues and potential resistance development. In contrast, our top natural compounds exhibit docking scores ranging from –7.85 to –7.22 kcal/mol, indicating a stronger and more stable interaction with the FolP active site. This enhanced binding affinity suggests that our compounds may offer superior inhibitory potency compared to both traditional sulfonamides and previously reported pterin-based inhibitors. Given that improved binding energies often correlate with higher biochemical efficacy, our results support the potential of these natural ligands as promising next-generation FolP inhibitors with optimized pharmacological properties and reduced susceptibility to resistance mechanisms.

The Discover Studio client tool was utilized to analyze the molecular interactions between the chosen compounds and the binding pocket of the FolP receptor. The main types of interactions discovered were halogen, alkyl, pi-stacked amide, carbon–hydrogen bond, conventional hydrogen bond, and pi–sulfur interactions. The binding affinities and docking scores of each of the best candidate compounds are largely determined by these interactions. A detailed examination of the inhibitor–FolP complex reveals a multi-faceted stabilization mechanism. The key hydrogen bonds between the inhibitors and conserved active site residues such as ARG253, Tyr141, and more in Figure 3 serve to anchor the inhibitor firmly within the pterin-binding pocket. This hydrogen bonding network not only orients the inhibitor optimally but also reduces its conformational flexibility, thereby enhancing the binding affinity. In parallel, hydrophobic interactions, most notably π–π stacking between aromatic portions of the inhibitors and residues like PRO55, LYS213, and ARG214, complement these polar interactions by facilitating tight packing within the pocket. Furthermore, water-mediated bridges contribute additional stability by connecting distant polar groups on the inhibitor and enzyme, thereby compensating for any steric mismatches. The cooperative effect of these interactions creates a highly stable inhibitor–protein complex, which is critical for outcompeting the natural substrate, PABA. These structural insights not only rationalize the observed docking scores but also provide a solid foundation for the future optimization of these inhibitors to enhance potency and selectivity.

The ADMET properties and toxicity risks of the selected compounds were evaluated, and it was discovered that the predicted values of three compounds (LTS0262854, LTS0241035, and LTS0033598) were within an acceptable range. This step provides information about the pharmacokinetic properties, bioavailability, and potential side effects of the compounds. ADMET analysis assists in removing compounds with unfavorable properties and focusing on those with optimal drug-like properties [35,36]. Our ADMET analysis, performed using OSIRIS Property Explorer, reveals that our top natural compounds possess highly favorable pharmacokinetic profiles. Notably, these compounds have cLogP values less than 5, and TPSA that collectively predict good membrane permeability and oral bioavailability. The calculated log S values and overall drug scores further suggest efficient solubility and absorption. In addition, toxicity risk assessments indicate that our compounds have low potential for mutagenicity, tumorigenicity, irritation, and reproductive toxicity (Table 3). In contrast, established leprosy therapeutics, such as dapsone and sulfonamides, although effective are frequently limited by suboptimal solubility and adverse side effects during long-term administration. Given that leprosy treatment often requires prolonged drug use, our compounds’ improved ADMET profiles may lead to better patient compliance and lower incidences of adverse effects. Combined with their superior binding affinities (docking scores between –7.85 and –7.22 kcal/mol), these pharmacokinetic advantages underscore the potential of our natural ligands as next-generation FolP inhibitors with enhanced efficacy and safety profiles.

MD simulations were used to gain a better understanding of the dynamic behavior of FolP–compound interactions. MD simulations provide insights into the stability of the complexes over time, providing a dynamic perspective on the binding interactions [37]. This step is critical for identifying compounds that maintain stable binding across multiple conformations. These compounds were stable as potent inhibitors inside the protein binding pocket based on MD simulations.

While in silico studies can provide useful insights, it is critical to recognize their limitations, such as the simplification of complex biological systems. In order to confirm the inhibitory potential of the selected natural compounds against the FolP protein, additional experimental validation, including in vitro and in vivo studies, will be required.

5 Conclusions

This comprehensive in silico study represents a strategic approach in the early stages of leprosy drug discovery. We hope to identify natural compounds with the potential to serve as effective FolP protein inhibitors in M. leprae by combining molecular docking, ADMET prediction, and MD simulation. The success of this research could pave the way for the development of novel, safe, and effective anti-leprosy drugs, contributing to global efforts to combat this neglected tropical disease.