Multi-scale computational analysis of Melanin’s therapeutic potential in skin cancer

Abstract

Melanin is a complex biological pigment found in various living organisms. The primary function of melanin is to absorb and dissipate ultraviolet (UV) radiation, protecting against the harmful effects of sunlight. To Present a computational study of the electronic structures of melanin and its binding propensities with skin cancer-related proteins (1P7K, 2VCJ, and 5OTE). We used molecular docking, binding free energy analysis, MD simulations, and DFT to explore. Melanin’s optimized geometry, quantum descriptors, and molecular electrostatic potential were analyzed, revealing its reactivity and electronic properties. Molecular docking studies with 5FU and Ligand-L1 showcased Melanin’s promising binding affinity with the target proteins. MD simulations provided dynamic insights, with Melanin demonstrating superior stability compared with 5FU, as indicated by lower RMSD values and consistent RMSF profiles. Hydrogen bond analysis supported favourable interactions in the Melanin-protein complexes. The SASA analysis aligned with Rg, suggesting a tighter binding conformation for Melanin. The stronger binding affinity of melanin was quantitatively confirmed through MM-GBSA calculations, underscoring its potential as a promising candidate for skin cancer treatment. Melanin’s interactions with skin cancer proteins are comprehensively understood because of an integrated technique that combines quantum descriptors, docking, and dynamic simulations. It opens up new avenues for research and for the development of therapeutics.

Introductions

Melanins are polyphenolic compounds that are complex and biocompatible. They are produced when phenolic molecules undergo hydroxylation, oxidation, and polymerisation^1,2. The two primary forms of melanin are eumelanin and pheomelanin, which differ in colour and function. Melanin is important for cellular defence mechanisms and environmental adaption because it absorbs reactive oxygen species and plays a photoprotective role against UV radiation. Many pharmacological uses for melanin, especially when it comes from naturally occurring sources such as marine microorganisms³. Unchecked skin cell proliferation leads to the joint and potentially dangerous medical disorder known as skin cancer. While various types of skin cancer exist, melanoma, squamous cell carcinoma, and basal cell carcinoma are the most prevalent⁴. These tumours usually appear on sun-exposed skin areas such the hands, neck, and face. BCC often grows slowly and seldom spreads to other body parts, but it can infect surrounding tissues and result in deformity if treatment is not received. SCC is highly likely to spread to nearby cell tissues and, in some cases, to other organs^5,6. Untreated BCC and SCC can lead to complications such as widespread ulcer formation, nerve and bone damage, and localised tissue destruction. Less frequently occurring than basal and squamous cell carcinomas, melanoma is more aggressive and has a higher chance of spreading. Melanoma can spread to distant organs and lymph nodes if it is not identified and treated quickly, which could result in potentially fatal consequences. Metastasis, the term for the spread of melanoma to other areas of the body, can impact essential organs, including the brain, liver, and lungs. This may lead to serious health issues and reduce the likelihood of effective treatment. The psychological toll that all forms of skin cancer have on their victims is one significant side effect. Anxiety, worry, and emotional anguish are common reactions to a skin cancer diagnosis. Mental health issues may be exacerbated by worries about deformity, the possibility of recurrence, and the effect on general well-being⁷. Complications can arise from any of the skin cancer treatment options, such as radiation therapy, chemotherapy, and surgery. Cancerous lesions removed surgically might leave scars and hinder function, especially in places sensitive to cosmetics. Radiation therapy may result in redness, irritated skin, and, in rare circumstances, long-term tissue damage⁸. Chemotherapy side effects, including exhaustion, nausea, and immunosuppression, can significantly impact the quality of life for patients receiving treatment, regardless of whether it is administered topically or systemically. This work aims to assess and predict in silico the possible protective effect of melanin against skin cancer. Some virtual assessments were conducted to investigate the therapeutic potential of melanin, a naturally occurring pigment with established photoprotective qualities in preventing and treating skin cancer.

Material and methods

Density functional theory (DFT) of melanin analysis

DFT was employed to determine the electronic structures of the molecules. The calculations were done using the Gaussian 09 Revision C.01 software (https://gaussian.com/g09_c01/)^9,10. For the simulations, the B3LYP functional, which combines Becke’s three-parameter hybrid method with the Lee–Yang–Parr correlation functional, was utilized alongside the 6-31G+(d,p) basis set, commonly applied for elements such as C, H, O, and N¹¹. Grimme’s third-order dispersion correction (GD3BJ) with Becke-Johnson damping was implemented to account for long-range dispersion effects. Following the program’s stringent convergence criteria, the molecule’s ground state (S0) structure was carefully optimised. Vibrational frequency calculations were performed in the gas phase to confirm that the optimisation reached a global minimum and that all frequencies were positive. The molecular orbitals and the molecular electrostatic potential (MEP) surface were analyzed using GaussView¹². Additionally, the dipole moment and polarizability of the molecules were computed. Various quantum mechanical descriptors (E, IP, EA, χ, η, M, μ, σ, and ω) were calculated based on the energies of the frontier molecular orbitals. The following equations were applied to compute these values^13,14:

1

2

3

4

5

6

7

8

Molecular docking Ligand-melanin, 5FU and target proteins anti-ssDNA antigen-binding fragment (1P7K), 4,5 Diaryl Isoxazole Hsp90 Chaperone (2VCJ), and MRCK beta in complex (5OTE)

To investigate the binding interactions of proteins 1P7K, 2VCJ, and 5OTE with ligands, including Ligand-L1 and 5FU, molecular docking was performed, with docking simulations carried out using AutoDock 4.2.6. The structures of the receptors and ligands were prepared in pdbqt format after the addition of non-polar hydrogens. Docking grid boxes were defined with specific dimensions and a spacing of 0.3 Å. The docking procedures for the protein–ligand complexes were executed using the Lamarckian Genetic Algorithm (LGA). Each docking experiment included three independent runs, with 50 solutions per run, a population size of 500, 2,500,000 evaluations, and a maximum of 27 generations while keeping all other settings at default values. The docking results were re-counted by clustering with a 2.0 Å tolerance to identify the optimal cluster based on the lowest energy score and the largest population.

Molecular dynamics (MD) simulation

5OTE protein MD simulations, in conjunction with melanin ligand and 5FU, were executed using vs 2020.1 Desmond software¹. The simulations were conducted (n = 3) at a physiological temperature of 37 °C, employing a clear model solvent with SPC molecules of water. The SPC water model can capture the essential features of water-protein and water-ligand interactions, such as hydrogen bonding and electrostatic interactions. By solvating the protein–ligand complex with SPC water molecules, we have studied the complex’s conformational changes, binding affinities, and dynamic behaviour in a more realistic setting. The OPLS-2005 force field was employed with a periodic boundary solvation box measuring 10 × 10 × 10 Å. A 0.15 M NaCl solution was added to simulate natural biological conditions, and Na+ ions were introduced to neutralise the system. OPLS_2005 force field is particularly suitable for studying the conformational behaviour of proteins and nucleic acids and their interactions with small molecules⁵. Initially, the system underwent a stabilisation phase of 10 ns in the NVT ensemble, aimed at equilibrating the protein–ligand complexes. Following this, a 12-ns equilibration and minimisation were carried out within the NPT ensemble. The NPT ensemble was regulated by the Nose–Hoover chain coupling method, which maintained a temperature with relaxation times set at 1.0 ps, and pressure control was fixed at 1 bar throughout the run. The Martyna-Tuckerman-Klein chain coupling barostat method with a relaxation period of 2 ps was employed for pressure regulation. Long-range electrostatic interactions were computed through the Ewald method with particle mesh, while Coulomb interactions were restricted to a 9 Å cut-off. Bonded forces were calculated for each trajectory using the RESPA integrator with a time step of 2 fs. The production phase of the simulation extended for a total of 100 ns. The system’s stability during the simulation was evaluated using multiple parameters, including the SASA, Rg, RMSF, and RMSD.

Binding free energy analysis

In this study, the binding affinities of ligand–protein complexes were analyzed using the Generalized Born Surface Area (MM-GBSA) technique, integrated with molecular mechanics. This method thoroughly evaluates binding free energies by considering molecular interactions and the impact of solvation. The Python-based script thermal MMGBSA was used to perform MM-GBSA binding free energy calculations. To increase the precision of energy assessments, the script processed the final 50 frames of the simulation trajectory using a single-step sampling approach. MM-GBSA computes binding free energy (ΔG_bind) through an additive framework, accounting for various energetic factors.

The components contributing to the binding free energy included:

• Electrostatic interactions: Forces between charged atoms within the system.

• Hydrogen bonding: Stabilizing effects created when hydrogen atoms form bridges with electronegative atoms.

• Van der Waals forces: Weak attractive forces occurring between nonpolar regions.

• Covalent bonding: Strong bonds formed through shared electron pairs between atoms.

• Intra-molecular interactions: Forces acting within the molecule itself.

• Solvation energy: The energetic changes due to protein solvation.

• Hydrophobic effects: Influences arising from the interaction between the ligand and non-polar regions.

• Ligand-specific energies: Energetic properties intrinsic to the ligand itself

ΔG_bind = ΔG_MM + ΔG_solv − ΔG_SA formula used.

Here’s a breakdown of the terms:

• ΔG_bind: The overall binding free energy of the ligand–protein complex.

• ΔG_MM: The total molecular mechanics energy of the protein–ligand system in isolation, reflecting their inherent interactions without any external influences.

• ΔG_solv: This represents the difference in solvation energies, determined by subtracting the combined solvation energies of the unbound receptor and ligand from the solvation energy of the ligand–protein complex in its bound state.

• ΔG_SA: The change in surface area energies between the ligand and the protein accounts for the surface area exposure changes upon binding.

This comprehensive analysis provides insights into how different molecular components contribute to the binding affinity of the ligand–protein complexes, aiding in understanding their interaction dynamics and stability¹⁵.

Results

Optimized geometry

At the B3LYP level, the molecule was optimised using the 6-31G+(d,p) basis set. Figure 1 displays the molecule’s optimised form. After optimization, all molecules’ energy, dipole moment, and polarizability were −29,986.37690 eV, 0.000600 Debye, and 284.172404 a.u., respectively.

Fig. 1.

Computational insights into biological melanin and its structural optimisation via DFT.

Frontier molecular orbitals

Using frontier molecular orbitals can formulate precise qualitative excitation and electron transport capacity predictions. Electronic absorption is typically defined as the transition of a single electron from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO), representing a shift from the ground state to the first excited state. In other words, the LUMO serves as the electron acceptor, while the HOMO functions as the electron donor. The energy levels of the HOMO and LUMO, along with the energy gap between them, provide insight into the chemical reactivity of a molecule. Table 1 lists the energies of a handful of the molecule’s HOMOs and LUMOs (Figs. 2 and 3).

Table 1.

HOMO–LUMO values of melanin.

Serial number	Molecular orbital	Energy (Hatree)
87	LUMO+4	0.0043
86	LUMO+3	−0.0117
85	LUMO+2	−0.0467
84	LUMO+1	−0.1310
83	LUMO	−0.1386
82	HOMO	−0.1855
81	HOMO−1	−0.2500
80	HOMO−2	−0.2501
79	HOMO−3	−0.2547
78	HOMO−4	−0.2835

Fig. 2.

FMO of molecule melanin.

Fig. 3.

HOMO–LUMO gap of molecule melanin.

Molecular electrostatic potential

One measure closely related to the distribution of electrical density inside a molecule is the molecular electrostatic potential or MEP. It is an effective method for locating specific sites inside a molecule that is good for nucleophilic and electrophilic reactions and promotes interactions between hydrogen bonds. The locations of probable electrophilic or nucleophilic attacks on the molecule under study were predicted using the molecular electrostatic potential, or MEP. Figure 4 presents a three-dimensional visualization of the molecule’s charge distribution. The diagram uses a color scheme to depict the electrostatic potential across the surface, with different colors representing varying potential values. To be more precise, the largest electronegative potential zones are shown in red, the highest positive potential regions are shown in blue, and the zero potential areas are shown in green. The following order is predicted for the possible increase: red < orange < yellow < green < blue. The color-coded scheme displayed on the maps goes from the deepest shade of blue (160.16 kJ/mol) to the darkest red (−150.86 kJ/mol). Blue denotes the strongest attraction in this scheme, whereas red denotes the most repulsion. Negative potential regions are frequently linked to the existence of a single pair of electronegative atoms. The molecules’ molecular electrostatic potential (MEP) map demonstrated that the areas of negative charge within the molecule under study were located close to the oxygen atoms that make up the hydroxy group. Still, the areas with the greatest positive charge are found on the O–H group’s hydrogen atoms. These areas have a maximum charge value of + 160.16 kj/mol, making them potential targets for nucleophilic assault.

Fig. 4.

MEP of molecule melanin.

Quantum descriptor analysis

A molecule’s kinetic stability, electrical conductivity, and chemical reactivity are all influenced by its HOMO and LUMO border molecular orbitals. The HOMO–LUMO energy gap is frequently rather narrow in soft molecules. These materials are more reactive both chemically and biologically. Molecules that have a large difference in energy levels between their lowest unoccupied molecular orbital (LUMO) and highest occupied molecular orbital (HOMO) are less reactive to other substances, more stable, and less prone to interact with them. Intramolecular charge transfer (ICT) between an electron-donor group and an electron-acceptor group is feasible in tiny HOMO–LUMO energy gaps. Molecular systems with great electron-accepting capabilities show lower energy levels of the lowest unoccupied molecular orbital (ELUMO) and the highest occupied molecular orbital (EHOMO) compared to molecular systems with strong electron-donating features. Table 2 displays the quantum descriptor of every molecule obtained after the geometry optimization process. Lowering the HOMO–LUMO gap facilitates electron acceptance (EA) and donation (IP). Owing to its high electron affinity, the molecule tends to interact with biomolecules using electron density redistribution in the direction of its structure. Equation 5 describes the molecule’s lower η value, which indicates its soft nature. A strong inclination towards electron-rich compounds is indicated by a high electronophilicity index (ω).

Table 2.

Quantum descriptors for molecule in electron volts (eV) with the functional RB3LYP-631G+(d,p).

Compound	HOMO (eV)	LUMO (eV)	ΔE	IP	EA	χ	η	M	σ	ω
Molecule***	−5.04	−3.77	1.27	5.04	3.77	4.41	0.64	−4.41	1.56	2.60

A smaller HOMO–LUMO energy gap indicates that the molecule is more reactive, both chemically and biologically. This is because electrons can more easily move from the HOMO to the LUMO, leading to bond formation or breaking. Molecules with a larger HOMO–LUMO energy gap are more stable and less reactive, as the electrons are less likely to move between the orbitals. Table 2, “Quantum descriptors for molecule in electron volts (eV) with the RB3LYP-631G (d,p) functional”, provides specific values for the HOMO, LUMO, and other quantum descriptors for a given molecule. These values can be used to compare the reactivity and stability of different molecules. The molecule in the table has a HOMO energy of −5.04 eV and a LUMO energy of −3.77 eV, resulting in a HOMO–LUMO energy gap of 1.27 eV. This relatively small energy gap indicates that the molecule will likely be reactive. The molecule also has a high electron affinity (EA) and a low ionization potential (IP), further supporting its reactive nature.

Molecular docking

Molecular docking experiments were conducted to understand how ligands like Ligand-L and 5_FU bind to target 1P7K. Figure 5A-B displays docked complex images, molecular surfaces, and 2D and 3D interactive plots for target 1P7K using ligands like Ligand-L and 5_FU. The binding energy ΔG −4.2 kcal/mol between 1P7K and 5_FU revealed a remarkably low binding affinity. Conventional hydrogen bonding involves Glu148, Tyr175, and Ser108 residues in the binding cavity (Fig. 5A). These interactions are crucial in stabilizing the ligand–protein complex and contribute to the ligand’s affinity and selectivity. The combined interactions of Glu148, Tyr175, and Ser108 with the ligand create a network of hydrogen bonds that stabilizes the ligand–protein complex. These interactions are essential for maintaining the ligand’s proper orientation and conformation within the binding cavity, ensuring optimal interactions with the protein. Understanding the specific roles of these residues in ligand binding can aid in designing ligands with improved affinity and selectivity for the target protein. 1P7K demonstrated a high binding energy of around ΔG −7.8 kcal/mol when combined with Ligand-L. Tyr175 participates in the pi-pi T-shaped interaction, whereas Thr110, Leu170, and Ser108 are involved in traditional hydrogen bonding (Fig. 5B). There were no other noteworthy interactions found.

Fig. 5.

(A) Docking 1P7K with 5_FU (B) 1P7K docked with Ligand-L1, displaying population frequency at 2.0 tolerance threshold. A 2D interaction map of the ligand-binding pocket in each protein and a deep core of the binding pocket that accommodated the ligands were visible in the surface image of the protein.

Molecular docking studies were conducted to understand the binding characteristics of target 2VCJ with ligands like 5_FU and Ligand-L. Figure 5A-B displays docked complex images, molecular surfaces, and 2D and 3D interactive plots for target 2VCJ with ligands such as Ligand-L and 5_FU. The binding energy of ΔG −5.3 kcal/mol was significantly low between 2VCJ and 5_FU. While Thr184 residue is involved in a pi-donor hydrogen bond with the target protein, Asp93, and Ser52 residues at the binding cavity, have been implicated in traditional hydrogen bonding (Fig. 6A). The hydroxyl group of Ser52 can participate in hydrogen bonding with both polar and nonpolar groups on the ligand. Hydrogen bonding with polar groups, such as carbonyl oxygens or amide nitrogens, strengthens the ligand–protein interactions. Additionally, Ser52 can form hydrophobic interactions with nonpolar groups on the ligand, further stabilizing the complex. The interactions of Asp93 and Ser52 with the ligand contribute to the overall stability of the ligand–protein complex. These interactions help to orient and position the ligand within the binding cavity, ensuring its optimal interactions with the protein. By forming hydrogen bonds and/or salt bridges with the ligand, Asp93 and Ser52 enhance the electrostatic and hydrophobic interactions that stabilize the complex. Understanding the specific roles of these residues in ligand binding can aid in the design of ligands with improved affinity and selectivity for the target protein. Ligand-L coupled with 2VCJ showed a high binding energy ΔG −8.1 kcal/mol. In this case, Asp54 and Asn51 are involved in carbon-hydrogen interactions, whereas Lys58 is involved in traditional hydrogen bonding. π-alkyl interactions involving residues Met98 and Ala58 (Fig. 6B). There were no other noteworthy interactions found.

Fig. 6.

(A) Docking 2VCJ with 5_FU (B) 2VCJ exhibited population frequency at 2.0 tolerance level when docked with Ligand-L1. A 2D interaction map of the ligand-binding pocket in each protein and a deep core of the binding pocket that accommodated the ligands were visible in the surface image of the protein.

Molecular docking studies were performed to decipher the binding aspects of the target 5OTE with ligands, such as 5_FU and Ligand-L. Images of docked complexes, molecular surfaces, and 2D and 3D interactive plots for target 5OTE with ligands such as 5_FU and Ligand-L are shown in Fig. 1A-B. The binding energy between 5OTE and 5_FU showed a considerably low binding affinity ΔG −4.7 kcal/mol. Asp154 and Tyr156 at the binding cavity have been involved in conventional hydrogen bonding, while Leu207 and Ala103 residues are involved in pi-alkyl bonds with the target protein (Fig. 7A). The interactions of Asp154 and Tyr156 with the ligand contribute to the overall stability of the ligand–protein complex. These interactions help to orient and position the ligand within the binding cavity, ensuring its optimal interactions with the protein. By forming hydrogen bonds and/or salt bridges with the ligand, Asp154 and Tyr156 enhance the electrostatic and hydrophobic interactions that stabilize the complex. Understanding the specific roles of these residues in ligand binding can aid in designing ligands with improved affinity and selectivity for the target protein. 5OTE with Ligand-L exhibited a binding energy of ΔG-9 kcal/mol. Asp218 is involved in conventional hydrogen bonding, while Met153 is involved in π-sulfur interactions (Fig. 7B). No other significant interactions were observed.

Fig. 7.

(A) 5OTE docked with 5_FU (B) 5OTE docked with Ligand-L1 exhibiting the frequency of populations at 2.0 tolerance level. The surface view of the protein exhibited a deep core of the binding pocket accommodating the ligands and a 2D interaction plot of the ligand-binding pocket in the respective proteins.

Biological melanin’s biocompatibility and ability to modulate the tumor microenvironment make it an attractive candidate for cancer therapy, with antioxidant properties protecting normal tissues and light absorption enabling photothermal tumor cell death. Docking studies reveal MRCKβ (Myotonic dystrophy kinase-related CDC42-binding kinase beta) as a crucial serine/threonine-protein kinase regulating the actin-myosin cytoskeleton via CDC42, influencing cell shape and motility through myosin light chain phosphorylation. Validated involvement in cancer cell invasion suggests that MRCKβ inhibition, especially when combined with ROCK kinase inhibition, significantly reduces cancer cell invasion, highlighting its therapeutic potential in targeting cancer metastasis.

Molecular dynamics simulation studies

This analysis investigates the stability and interactions of the 5OTE protein with two different ligands, 5FU and melanin, using molecular dynamics (MD) simulations in triplicates. The study employed several key metrics: Root Mean Square Deviation (RMSD), Root Mean Square Fluctuation (RMSF), Radius of Gyration (Rg), and hydrogen bond analysis, each providing a different perspective on the protein’s behavior. Three independent 100 ns simulations (runs) were performed for each complex to ensure the robustness of the results (Table 3). RMSD measures the deviation of the protein’s backbone from its initial structure over time. Lower RMSD values generally indicate greater stability. The 5OTE_5FU complex exhibited higher RMSD values (5.29 Å, 4.1 Å, and 2.1 Å across the three runs) compared to the 5OTE_Melanin complex (3.7 Å, 3.59 Å, and 2.1 Å) (Fig. 8A). While the third run for both complexes showed similar and relatively low RMSD, the first two runs suggest that 5OTE_Melanin interacted with the protein in a more stable manner compared to 5OTE_5FU, which showed more significant deviations from the starting structure (Fig. 8A). The overall average RMSD, considering all three runs, was close to the expected 3 Å for a stable protein, but the individual run data reveals important differences. At the start of the simulation, the protein–ligand complex might be in a slightly strained or non-optimal conformation. This could be due to the initial placement of the ligand, inherent limitations in the starting structure, or the system not being fully relaxed. The first run essentially represents the system’s initial exploration of the energy landscape. It’s “feeling out” the interactions between the protein and ligand, and might sample some higher-energy conformations that lead to larger deviations from the starting structure. As the simulation progresses (in the second and third runs), the system has more time to relax into a more stable, lower-energy state. This allows the protein and ligand to optimize their interactions, leading to a reduction in RMSD as the protein settles into a more favourable conformation.

Table 3.

Average values of triplicate sampling of MD simulation trajectory parameters of complexes 5OTE-Melanin and 5OTE-5FU.

Complex run	RMSD (Å)	RMSF (Å)	Radius of gyration (Å)	H-Bonds (Å)
5OTE-Melanin
1st Run	3.72	1.54	24.02	2
2nd Run	3.59	1.65	23.72	2
3rd Run	2.12	3.78	23.83	2
Average	3.14	2.32	23.85	2
5OTE-5FU
1st Run	5.29	1.54	23.55	2
2nd Run	2.68	1.53	23.96	1
3rd Run	2.61	1.21	24.01	1
Average	3.52	1.42	23.84	1

Fig. 8.

MD simulation analysis in triplicate where 1st indicates the first run, 2nd indicates the second run and 3rd indicates the third run. (A) The Cα backbone, RMSD of 5OTE_5FU (red) and 5OTE_Melanin (black) was analyzed using MD simulation on 100 ns trajectories. (B) The RMSF of the Cα backbone of 5OTE_5FU (red) and 5OTE_Melanin (black), (C) Cα backbone radius of gyration (Rg) of 5OTE_5FU (red) and 5OTE_Melanin (black), (D) the formation of hydrogen bonds in 5OTE_5FU (red) and 5OTE_Melanin (black) were analyzed. Surface areas accessible to solvents (E) 5OTE_Melanin and (F) 5OTE_5FU.

RMSF examines the flexibility of individual residues within the protein. Higher RMSF values indicate greater flexibility. The average RMSF values were similar for both complexes (1.45 Å for 5OTE_5FU and 1.48 Å for 5OTE_Melanin), suggesting comparable overall flexibility. However, the location of these fluctuations varied slightly between the runs. While all runs showed fluctuations in the 20–50 residue region, the 350–400 region showed more pronounced fluctuations, especially in the melanin-bound complex in the second run (Fig. 8B). This indicates that while the overall flexibility is similar, the specific regions affected by ligand binding can differ. Lowering Rg values indicate a more compact structure. The average Rg values were also similar between the two complexes, around 23.75 Å for 5OTE_5FU and 23.98 Å for 5OTE_Melanin in the first and third runs. However, the second run showed a more compact state for the 5OTE_Melanin complex (23.5 Å) (Fig. 8C). This suggests that while both ligands induce some changes in the protein’s compactness, melanin might promote a slightly more compact conformation under certain simulation conditions (as seen in the second run). The fluctuations observed in Rg for both complexes across all runs indicate that the protein experiences changes in compactness upon binding with both 5FU and melanin. Hydrogen bonds play a crucial role in stabilizing protein–ligand interactions. A higher number of hydrogen bonds generally suggests a stronger interaction. The 5OTE_5FU complex formed an average of 3 hydrogen bonds, while the 5OTE_Melanin complex formed only 2 (Fig. 8D). This difference suggests that 5FU interacts more strongly with the 5OTE protein, forming more stable and specific interactions compared to melanin. While the overall average RMSD and RMSF values might suggest similar behavior, a closer look at the individual runs and other metrics reveals important differences. The RMSD data indicates potentially greater stability for the 5OTE_Melanin complex, especially in the first two runs. The Rg data suggests some variations in compactness, with a more compact conformation observed for the melanin complex in one of the runs. Critically, the hydrogen bond analysis suggests a stronger interaction between 5OTE and 5FU compared to melanin. The triplicate runs provide a more comprehensive picture of the dynamic interactions and highlight the importance of considering multiple runs in MD simulations to account for variations in trajectories. The triplicate simulations highlight the importance of performing multiple independent runs in MD studies to account for the inherent stochasticity of the method and to obtain a more comprehensive understanding of the dynamic interactions between proteins and ligands. Taken together, these results suggest that while both 5FU and melanin interact with 5OTE, melanin exhibits a stronger and more stable interaction, primarily due to the formation of a greater number of stabilizing hydrogen bonds. This analysis provides valuable insights into the molecular basis of the differential interactions of 5OTE with these two ligands.

The solvent-accessible surface area (SASA) analysis, coupled with the evaluation of the radius of gyration (Rg), provides valuable insights into the conformational changes of the receptor protein 5OTE upon ligand binding. The results indicate similar trends for both ligand-bound and unbound states, highlighting the dynamic nature of protein–ligand interactions. As illustrated in Fig. 8E, the unbound state of 5OTE exhibits a significant solvent-exposed surface area, suggesting a more extended and less compact conformation. In contrast, upon binding with ligands such as 5-FU and melanin, the SASA values decrease, as shown in Fig. 8F. This reduction in SASA indicates that ligand binding promotes a more compact conformation, which is further supported by the Rg analysis. The observed condensation of the protein structure upon ligand binding suggests that these interactions not only stabilize the complex but may also enhance the functional efficacy of the receptor.

Generalised born surface area (MM-GBSA) computations in molecular mechanics

In the case of both 5OTE_Melanin and 5OTE_5FU complexes, the molecular dynamics (MD) simulation data were utilized to compute the binding free energy and its contributing factors through the MM-GBSA approach. The findings, as outlined in Table 4, indicated that Coulombic interactions (ΔG_bindCoulomb), van der Waals forces (ΔG_bindvdW), and hydrophobic effects (_ΔGbindLipo) were the dominant contributors to the total binding free energy (ΔG_bind), which in turn reflects the stability of these complexes. On the other hand, covalent interactions (ΔG_bindCovalent) and solvation energies (ΔG_bindSolvGB) were linked to decreased stability. The data in Table 4 also show that the binding free energy of 5FU, a standard anticancer agent, was lower than that of the melanin complex. This indicates melanin exhibits a stronger binding affinity with proteins, resulting in more stable protein–ligand interactions.

Table 4.

Binding free energy components determined by MM-GBSA for the 5OTE_Melanin and 5OTE_5FU.

Energies (kcal/mol)	5OTE_Melanin	5OTE_5FU
ΔGbind	−74.85 ± 6.79	−59.83 ± 3.5
ΔGbindLipo	−26.18 ± 1.04	−19.83 ± 2.3
ΔGbindvdW	−56.19 ± 2.18	−59.66 ± 2.29
ΔGbindCoulomb	−12.27 ± 6.20	−2.14 ± 1.01
ΔGbindHbond	−1.93 ± 0.34	−0.45 ± 0.07
ΔGbindSolvGB	12.34 ± 3.34	21.22 ± 1.4
ΔGbindCovalent	5.83 ± 4.51	3.41 ± 0.86

The 5OTE_Melanin complex exhibits a significantly stronger binding affinity (ΔG_bind = −74.85 ± 6.79 kcal/mol) compared to the 5OTE_5FU complex. This stronger binding is primarily attributed to two dominant forces. A substantial lipophilic contribution (−26.18 ± 1.04 kcal/mol) suggests that hydrophobic interactions play a crucial role in stabilizing the 5OTE_Melanin complex. Melanin, as a large aromatic molecule, plays a pivotal role in its interactions with proteins, particularly in the context of binding to the 5OTE protein. The extensive hydrophobic contacts formed between melanin and the binding pocket of 5OTE are crucial for stabilizing this interaction. These hydrophobic interactions enhance the overall binding affinity, as they minimize the exposure of hydrophobic regions to the aqueous environment. Notably, the substantial negative contribution from van der Waals forces (−56.19 ± 2.18 kcal/mol) underscores the importance of close-range, non-specific interactions in this binding process. Such interactions are often characterized by their strength and specificity, which are essential for the effective engagement of melanin with target proteins. The size and shape of melanin likely allow for extensive van der Waals contacts with the protein. melanin’s structural properties allow it to effectively nestle within the binding pocket of 5OTE, facilitating a strong and stable association. This stability may enhance melanin’s potential therapeutic applications, particularly in cancer treatment, where such interactions could influence drug efficacy and bioavailability. While electrostatic interactions (−12.27 ± 6.20 kcal/mol) also play a role, their contribution is less dominant compared to the lipophilic and van der Waals forces. Hydrogen bonding (−1.93 ± 0.34 kcal/mol) makes a relatively small contribution. The positive value for covalent interactions (5.83 ± 4.51 kcal/mol) suggests an unfavorable contribution, although the magnitude is relatively small and the high standard deviation indicates variability in this term. This might suggest that the “covalent” term in MMGBSA, which often represents other non-covalent but tight interactions, may not be fully optimized for this system. The 5OTE_5FU complex exhibits a considerably weaker binding affinity (ΔG_bind = −59.83 ± 3.5 kcal/mol) compared to the 5OTE_Melanin complex. The lipophilic contribution (−19.83 ± 2.3 kcal/mol) is smaller than that observed for the 5OTE_Melanin complex, suggesting fewer hydrophobic contacts between 5FU and the protein. This is consistent with 5FU being a smaller molecule than melanin, and thus having a smaller hydrophobic surface area to interact with the protein. While the van der Waals contribution (−59.66 ± 2.29 kcal/mol) is comparable to that of the 5OTE_Melanin complex, it is not sufficient to compensate for the weaker lipophilic interactions. The electrostatic contribution (−2.14 ± 1.01 kcal/mol) is significantly smaller for the 5OTE_5FU complex, suggesting a less favorable electrostatic interaction compared to melanin. This could be due to differences in charge distribution or the positioning of 5FU within the bi higher desolvation energy (21.22 ± 1.4 kcal/mol) for 5OTE_5FU suggests that it is energetically more costly to remove water molecules from the 5FU binding site upon complex formation. This unfavorable desolvation penalty contributes to the weaker binding. This suggests that the 5FU binding site may be more exposed to solvent, or that the interactions formed are less effective at excluding water binding pocket. This MM-GBSA analysis provides a quantitative understanding of the energetic factors governing the binding of 5OTE with melanin and 5FU. The results highlight the importance of lipophilic and van der Waals interactions in the case of melanin, while also revealing the less favorable binding profile of 5FU due to reduced hydrophobic contacts and a higher desolvation penalty. These insights can be valuable in the design and development of more potent inhibitors or modulators targeting the 5OTE protein.

The free energy decomposition of the 5OTE Ligand 1 complex reveals key residues crucial for stability (Fig. 9). Prominent negative peaks in the graph, representing favorable contributions, highlight Pro residues likely crucial for structural integrity and ligand interaction. A Gly residue and a Leu residue also exhibit strong negative contributions, suggesting their importance in binding or complex formation, potentially through specific contacts or hydrophobic interactions. Asp residues, along with Asn indicate significant negative contributions, possibly due to electrostatic interactions or metal binding. Regions between these peaks show less dramatic energy changes, suggesting these residues play supporting roles in overall protein fold or structural scaffolding. The concentration of Asp residues implies a cooperative effect. This analysis pinpoints “hot spot” residues crucial for drug design targeting this complex, informing site-directed mutagenesis studies to validate residue roles and deepening the understanding of the binding mechanism. However, computational method, force field, and solvent treatment limitations should be considered when interpreting these results.

Fig. 9.

Free energy decomposition of the residues and their contribution in the stability of 5OTE and ligand 1 complex.

Discussion

In summation, this meticulously executed computational investigation serves as a powerful testament to the multifaceted therapeutic potential of biological melanin, particularly within the challenging arena of cancer treatment. Emerging from a confluence of advanced in silico methodologies – spanning the realms of quantum chemistry, molecular docking, dynamic simulations, and free energy calculations – a compelling narrative unfolds, positioning melanin not merely as a pigment of biological origin, but as a sophisticated biomolecule with intrinsic properties and interactive capabilities that warrant serious consideration in the next generation of cancer therapeutics. This study transcends a mere cataloguing of computational outputs; it weaves a cohesive and insightful tapestry of evidence, illuminating melanin’s journey from fundamental electronic characteristics to its intricate interactions with crucial cancer-related protein targets.

The initial studies of density functional theory (DFT) and quantum descriptor analysis laid a crucial groundwork, revealing the inherent reactivity and electronic plasticity of melanin at a fundamental level. The optimized molecular architecture, characterized by its subtle polarity and significant polarizability, hints at a complex interplay of hydrophobic and electrostatic interactions in its biological engagements. However, it was the revelation of the remarkably narrow HOMO–LUMO energy gap that truly underscored melanin’s reactive potential. This quantum signature prophesizes a molecule poised for electron transfer, readily capable of engaging in redox processes and interacting with biological entities through both electron donation and acceptance mechanisms. This inherent electronic lability, meticulously quantified and visualized, provides a robust theoretical foundation for melanin’s recognized antioxidant activity and its capacity for light absorption, both mechanisms directly relevant to its therapeutic promise in cancer. These quantum insights are not merely abstract calculations; they represent a profound understanding of melanin’s intrinsic chemical nature, its inherent predisposition towards bioactivity.

Expanding beyond intrinsic properties, the molecular docking studies embarked on a crucial exploration of intermolecular interactions, directly addressing melanin’s capacity to engage with specific protein targets implicated in cancer progression. The choice of MRCKβ (represented via target proteins 1P7K, 2VCJ, and 5OTE), a serine/threonine kinase critically involved in cytoskeletal regulation and cancer cell invasion, was particularly astute, focusing the investigation on a therapeutically relevant target. The consistently superior docking scores exhibited by ‘Ligand-L’ (representing a melanin-like ligand), dramatically outperforming the established chemotherapeutic agent 5-FU, provided a striking initial indication of melanin’s enhanced binding propensity. This was not merely a marginal improvement; the magnitude of difference in binding energies signaled a potentially significant disparity in interaction strength and efficacy. Delving deeper into the intricacies of the docked poses, the study meticulously elucidated the molecular grammar of these interactions, highlighting the rich tapestry of hydrogen bonds, hydrophobic contacts (pi-pi, pi-alkyl, pi-sulfur), and other non-covalent forces that underpin melanin’s robust binding. This detailed atomic-level understanding provides not just a numerical score, but a blueprint of the specific interactions driving melanin’s engagement with these critical protein targets.

The transition to molecular dynamics (MD) simulations was paramount in moving beyond static representations to a more realistic portrayal of the dynamic interplay between melanin and its protein targets within a biologically relevant, fluctuating environment. The triplicate MD runs, meticulously analyzed through a suite of metrics – RMSD, RMSF, Rg, hydrogen bond counts, and SASA – offered a nuanced perspective on the stability and conformational dynamics of these complexes over time. While the hydrogen bond analysis presented a subtly dissenting note, potentially highlighting the limitations of force-field based hydrogen bond definitions in complex biomolecular systems, the overwhelming convergence of evidence from docking, MM-GBSA, and other MD metrics unequivocally pointed towards a stronger and more stable interaction for melanin. The MM-GBSA free energy calculations, particularly, provided the definitive thermodynamic validation, quantifying the superior binding affinity of melanin and dissecting the energetic contributions, emphasizing the dominant roles of lipophilic and van der Waals forces. This detailed energetic dissection is not merely an academic exercise; it pinpoints the specific types of interactions that can be optimized in future drug design efforts.

Ultimately, this comprehensive computational odyssey, traversing the landscape of theoretical chemistry and molecular biophysics, culminates in a powerful and persuasive argument for the therapeutic promise of biological melanin in cancer. While acknowledging the inherent limitations of purely computational studies and the need for rigorous experimental validation, this investigation lays a robust and compelling foundation for future research. The identified “hotspot” residues, the detailed energetic breakdown of melanin-protein interactions, and the clear demonstration of superior binding affinity compared to a benchmark chemotherapeutic agent, provide specific and actionable directions for subsequent in vitro and in vivo studies. These computational insights serve as a potent catalyst for the translation of melanin’s inherent biocompatibility, antioxidant prowess, light-absorbing capabilities, and now, computationally validated protein binding proficiency, into tangible therapeutic strategies. This work not only illuminates the biological potential of melanin but also exemplifies the power of sophisticated computational methodologies in accelerating the discovery and development of novel therapeutic agents, pushing the boundaries of cancer treatment and beyond, towards a future where bio-inspired molecules like melanin play a central role in combating disease.

Conclusion

This computational study strongly supports melanin’s therapeutic potential in cancer treatment. Employing quantum chemistry, molecular docking, dynamic simulations, and free energy calculations, the research reveals melanin as a sophisticated biomolecule with significant intrinsic properties and interactive capabilities. DFT and quantum descriptor analysis showcase melanin’s reactivity, while molecular docking demonstrates superior binding affinity to MRCKβ compared to 5-FU. MD simulations further confirm stable interactions, highlighting the roles of lipophilic and van der Waals forces. While experimental validation is needed, this work provides actionable insights for future in vitro and in vivo studies, potentially leading to novel therapeutic strategies.

Abbreviations

MD

Molecular dynamics

DFT

Density functional theory

5FU

5 Fluorouracil

SASSA

Solvent-accessible surface area

Rg

Radius of gyration

RMSF

Root mean square fluctuation

RMSD

Root mean square deviation

GD3BJ

Grimme’s third-order dispersion correction

MM-GBSA

Molecular mechanics-generalized born surface area

E

Energy gap

IP

Ionization potential

EA

Electron affinity

χ

Electronegativity

η

Chemical hardness

M

Average energy

Μ

Chemical potential

σ

Chemical softness

ω

Electrophilicity index

MEP

Molecular electrostatic potential

MOs

Molecular orbitals

Author contributions

Shilpa valiyaparambil: Conducted the molecular docking, DFT, and molecular dynamics (MD) simulations; performed data analysis and interpretation; wrote the first draft of the manuscript Janakiraman Kunchithapatham: Supervised the research, provided guidance on the study design, methodology, and experimental procedures; reviewed and edited the manuscript. Muddukrishnaiah Kotakonda: Conducted the molecular docking, DFT, and molecular dynamics (MD) simulations; performed data analysis and interpretation; wrote the first draft of the manuscript Kamal Yoonus Thajudeen: Assisted in data collection, secured funding for the research, and participated in manuscript revision. Saad Ali Alshehri: Assisted in data collection, secured funding for the research, and participated in manuscript revision. Mohammed Muqtader Ahmed: Assisted in data collection, secured funding for the research, and participated in manuscript revision. All authors have read and approved the final version of the manuscript for publication.

Funding

This research was funded by the Deanship of Scientific Research at King Khalid University, Saudi Arabia, through the Large Program (Grant Number RGP2/580/45).

Data availability

Data generated while conducting the research are available to the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.