Integrated Computational Profiling Links Position-specific O-Methylation to Spontaneous Complexation and Improved Binding in Flavonoid

Abstract

Human epidermal growth factor receptor 2-positive (HER2+) breast cancer is characterized by amplification and overexpression of HER2. This study used a computational workflow to identify O-methylated flavonoids with favorable binding to the HER2 kinase domain and to clarify their binding modes. Among the screened compounds, Eupatolitin, Rhamnetin, Annulatin, Laricitrin, and Sterubin showed the best predicted affinities, with docking scores ranging from -10.092 to -9.260 kcal/mol. Molecular dynamics simulations were then used to assess complex stability, and density functional theory was applied to examine the reactivity of the lead compounds. These analyses were intended for computational prioritization, and experimental validation is required to confirm HER2 inhibition and biological activity. Eupatolitin, which bears methoxy groups at C-6 and C-7, showed the strongest predicted binding, suggesting that 6,7-O-methylation may favor HER2 recognition. This observation prompted a comparison of four parent flavonoids with their 6,7-O-methylated analogues. Across quercetin, myricetin, luteolin, and eriodictyol, 6,7-O-methylation consistently strengthened predicted HER2 binding: docking scores improved by 0.27-0.57 kcal/mol and MM/GBSA ΔGbind improved by 6-10 kcal/mol relative to the parent compounds. In contrast, natural O-methylation at other positions did not enhance binding. Overall, these findings identify Eupatolitin, Rhamnetin, Annulatin, Laricitrin, and Sterubin as promising computational HER2 hits and support position-specific O-methylation at C-6 and C-7 as a useful strategy for optimizing flavonoid scaffolds for HER2 targeting, but experimental validation is required.

Keywords

O-methylated flavonoids HER2 molecular docking molecular dynamics simulation DFT anticancer activity

1. Background

Breast cancer is the most commonly diagnosed cancer in women and remains a leading cause of cancer-related death worldwide.¹ Studies have projected that by 2030, the global number of new breast cancer cases diagnosed would reach 2.7 million yearly, with the number of deaths 0.87 million.²

Human Epidermal Growth factor Receptor 2 (HER2) positive breast cancer is a subset of breast cancer that is characterized by the amplification of the HER2 gene and/or protein overexpression of Human epidermal growth factor receptor 2.³ HER2-positive breast cancer is clinically characterized by HER2 gene amplification and/or HER2 protein overexpression.⁴ HER2 belongs to a specific family of receptor tyrosine kinase with other members of the family being Epidermal Growth Factor Receptor (EGFR) or HER1, HER3, and HER4.⁴ It is expected that up to 40–50% of patients with advanced HER2-positive breast cancer will eventually develop brain metastases.⁵

Although several HER2-targeted therapies are available, drug resistance, tumor recurrence, and disease progression remain major clinical challenges in HER2-positive breast cancer. Recent reviews continue to emphasize the need for new HER2-directed agents and improved therapeutic strategies to overcome acquired and intrinsic resistance.^4,6-8 It is therefore of interest to discover new drug candidates that can circumvent resistance and offer long-term clinical usage.

Flavonoids have been extensively studied and have become one of the most vital bioactives with the potential to treat cancer and oxidative stress, among others. More so, their special roles in pharmacology, dietary supplements, and cosmetics have been well documented.⁹ Structurally, they contain fifteen carbon atoms in their backbone, which are made up of three rings; they are labeled A,B, and C for easy identification.¹⁰

Structurally, O-methylated flavonoids are considered products of the post-modification of flavonoids as they are derived by the attachment of a methyl group to the oxygen of the hydroxyl terminal in the flavonoid skeleton.¹¹ The O-methylation positions of flavonoids are numerous due to the high number of hydroxyl groups in the flavonoid backbone.¹² Mono and dimethoxylated flavones have been found to show antiproliferative activity by their strong inhibitory activities against the transcription and activities of cytochrome p450 isoforms that activate carcinogenicity.^13,14

The O-methylation of one or more hydroxyl groups in flavonoids increases their transport across membranes, thereby aiding absorption.^15,16 In addition, previous studies have suggested that the methylated forms of some compounds have higher metabolic stability, are more orally bioavailable, and have increased bioactivity than the parent or unmethylated forms.¹⁷ To this end, this study aims to elucidate the potential anticancer activity of naturally occurring O-methylated flavonoids using computational techniques, including molecular docking, molecular dynamics simulation, and Density functional theory analysis. More importantly, the effect of natural methylation was studied by comparing the data obtained from the biomolecular interactions of the O-methylated flavonoids with their parent unmethylated forms.

2. Methods

2.1. Protein Targets Retrieval and Preparation

An integrated computational workflow was used to model the interactions between human epidermal growth factor receptor 2 (HER2) and the selected O-methylated flavonoids. The X-ray crystallographic structure of HER2 was downloaded from an online protein repository (rcsb.org) in PDB format and subsequently prepared for the intended molecular interaction. 3PP0 (the kinase domain of the receptor) was selected based on the presence, position, and binding of a co-crystallized ligand.¹⁸ The retrieved structure was incorporated into Maestro and prepared using the Protein Preparation Wizard script. The structure was initially preprocessed to identify problems such as missing side chains and overlapping atoms, among others. Water and non-standard bound moieties were deleted. The protein was then optimized and subsequently minimized using OPLS3 forcefield.

2.2. Ligand Mining and Preparation

The 2D structures of O-methylated flavonoids were retrieved from the PubChem repository in SDF format.¹⁹ The compounds were incorporated into Maestro and subsequently prepared using the Ligprep tool. Specifically, two-dimensional structures were modified to their corresponding three-dimensional structures. The structures were ionized at pH (7.2±0.2)

2.3. Grid Generation

A receptor grid maps the ligand-binding site of a protein, where a ligand binds and interacts with the protein.²⁰ The position and binding of the co-crystallized ligand of HER2 were employed to map the coordinates of the ligand-binding site. Using the Receptor Grid Generation tool on Maestro, the co-crystallized ligand was selected as prompted. This revealed the ligand-binding site and generated a grid box, which was subsequently used to guide the docking of the O-methylated flavonoids to HER2.

2.4. Molecular Docking

To validate the docking protocol, the co-crystallized ligand was extracted from the HER2 crystal structure and re-docked into the prepared binding site using the same grid-generation and docking settings applied to the test compounds. The accuracy of the protocol was assessed by calculating the root mean square deviation (RMSD) between the heavy-atom coordinates of the re-docked pose and those of the crystallographic ligand pose. An RMSD value of less than 2.0 Å was considered indicative of acceptable reproduction of the experimental binding mode and therefore supportive of the reliability of the docking procedure

The binding affinities of the test compounds toward HER2 were evaluated by Glide docking in Maestro after grid generation around the co-crystallized ligand-binding site. All 68 compounds were first screened using Glide standard precision (SP), with the receptor treated as rigid and the ligands treated as flexible. The top-ranked ligands were then re-docked using Glide extra precision (XP) to refine binding poses and ranking. Post-docking minimization was applied, and the best XP pose for each ligand was retained for interaction analysis and MM/GBSA rescoring. Unless otherwise stated, the docking scores reported are XP docking scores.²¹

2.5. MM/GBSA Calculation

Prime MM/GBSA in Maestro was used to rescore the docked HER2-ligand complexes obtained from the retained XP poses.^22,23 Binding free energy (ΔG_bind) was estimated for the optimized complex relative to the free receptor and free ligand using the OPLS3 force field, Prime rotamer search, and the VSGB 2.0 solvent model. The calculated MM/GBSA values were used for comparative ranking of ligand binding favorability rather than as direct estimates of experimental binding free energy, because MM/GBSA relies on simplified solvation models and approximate energy decomposition.²⁴

2.6. Molecular Dynamics Simulation

To further evaluate the stability of the complexes formed between the top-scoring compounds and the target receptor, molecular dynamics simulations were performed. The docked HER2-ligand complexes were exported in PDB format and simulated with GROMACS through the WebGRO server.²⁵ A standard WebGRO setup was used, with the GROMOS96 43a1 force field, SPC water model, and a triclinic simulation box. The systems were neutralized and maintained at 0.15 M NaCl by adding appropriate sodium and chloride ions. Energy minimization was carried out using the steepest descent algorithm for 5000 steps. The minimized systems were then equilibrated under NVT and NPT ensembles, using a commonly reported protocol of 100 ps NVT followed by 100 ps NPT, at 300 K and 1.0 bar. Production simulations were subsequently run for 100 ns using the leap-frog integrator. The resulting trajectories were analyzed in terms of root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), solvent-accessible surface area (SASA), and hydrogen-bond profiles using the simulation output files and XmGrace.²⁶ The MD simulation results via trajectory files were then analyzed using XmGrace software.²⁷

2.7. Density Functional Theory (DFT) and Theoretical Calculations

A density functional theory (DFT) analysis was carried out on the hit compounds to estimate the values of the frontier molecular orbitals using the Lee-Yang Parr exchange-correlation functional method (B3LYP) on Spartan 14.²⁸ The Structures of the compounds were built in Spartan and geometry optimized. For each compound, the lowest energy conformer was retained for quantum-chemical analysis. The values of the frontier molecular orbitals were obtained, namely, the energies of the highest occupied molecular orbital (E_HOMO) and the lowest unoccupied molecular orbital (E_LUMO). From the E_HOMO and E_LUMO values, theoretical calculations were carried out to obtain the values of other reactivity descriptors such as electronegativity, chemical hardness, chemical softness, and chemical potential.

2.8. Pharmacokinetic Profiling

The absorption, distribution, metabolism, excretion, and toxicity (ADMETox) descriptors of the top-scoring O-methylated flavonoids were predicted using SwissADME and Pro-Tox II.^29,30 The predicted descriptors include lipophilicity, water solubility, and permeability glycoprotein (P-gp) substrate candidacy among others. Per SwissADME predictions, consensus Log P, the average of five lipophilicity models (iLOGP, XLOGP3, WLOGP, MLOGP, and Silicos-IT Log P) was adopted as the lipophilicity of the test compounds. This property predicts how efficiently drugs or other substances are absorbed into cells and systemic circulation. The ESOL model of water solubility was used to predict the aqueous solubility of the O-methylated flavonoids.

3. Results

3.1. Molecular Docking

Sixty-eight naturally occurring O-methylated flavonoids were probed to identify their potential to bind to the kinase domain of HER2. Redocking the co-crystallized ligand reproduced the experimental binding pose with a heavy-atom RMSD of 1.80 Å, supporting the suitability of the docking protocol for pose prediction in the HER2 active site. Docking and MM/GBSA results are shown in Figure 1 (Figure 1A and B). About 45% of the selected compounds had docking scores ≤ -7.0 kcal/mol (Figure 1A). Thirty-four compounds returned docking scores in the range of -7.0 to -10.10 kcal/mol. The docking scores of the top 20 compounds ranged from -7.744 to -10.092 kcal/mol (Table 1). The five compounds selected for further investigation were not chosen solely on the basis of docking score, but also considering binding-pose quality, interaction with key HER2 active-site residues, consistency of favorable docking behavior, and suitability for further dynamic and energetic evaluation. Eupatolitin had the highest predicted binding affinity, with a docking score of -10.092 kcal/mol. Rhamnetin (-9.990 kcal/mol), Annulatin (-9.572 kcal/mol), Laricitrin (-9.569 kcal/mol), and Sterubin (-9.260 kcal/mol) followed. FDA-approved HER2 inhibitors used as reference compounds had the following docking scores: Neratinib (-8.782 kcal/mol), Afatinib (-7.120 kcal/mol), and Mubritinib (-4.252 kcal/mol).

Figure 1.

(A) Distribution of docking scores, (B) The MM/GBSA-based binding energy of the test compounds

Table 1.

Glide XP Docking Scores (kcal/mol) of the Top-Ranked O-Methylated Flavonoids and Reference HER2 Inhibitors

Methyl-flavonoids	Docking score (kcal/mol)
Eupatolitin	-10.092
Rhamnetin	-9.990
Annulatin	-9.572
Laricitrin	-9.569
Sterubin	-9.260
Nepetin	-9.132
Rhamnazin	-8.781
Meciadanol	-8.703
Azaleatin	-8.464
Hesperetin	-8.246
7-O-methylapigenin	-8.236
Ayanin	-8.222
Tricin	-8.193
7-O-methylnaringenin	-7.986
Sakuranetin	-7.986
Homoeriodictyol	-7.917
Peonidin	-7.844
Thevetiaflavone	-7.827
Poriol	-7.772
Psi-tectorigenin	-7.744
Neratinib	-8.782
Afatinib	-7.120
Mubritinib	-4.252

Bold type indicates the top five scoring compounds and the FDA-approved reference drugs included for comparison.

3.2. Specific Molecular Interactions Between the Top-Scoring Flavonoids and the Target Receptor

Figure 2 shows the binding interactions of the five lead flavonoids within the HER2 kinase domain. Eupatolitin interacted with Met801 residue on the kinase domain of HER2 by a hydrogen bond donor (2.25Ǻ) via the hydroxyl group on position 3 and a hydrogen bond acceptor (2.04 Ǻ) via the carbonyl group on position 4 (Figure 2A). Also, hydrogen bond donating contacts were made with Asp808 via the hydroxyl group on position 4’ (1.66 Ǻ) and position 5’ (1.69 Ǻ). Similarly, while binding to the ligand binding domain of the receptor, Rhamnetin formed two hydrogen bond contacts with Asp808 via the hydroxy group on position 3’ (1.63 Ǻ) and position 4’ (1.60 Ǻ) (Figure 2B). Another set of hydrogen bond contact was made with Met801; a hydrogen bond contact via the carbonyl group on position 4 (1.95 Ǻ) and a hydrogen bond donating contact with the hydroxyl group on position 3 (2.25Ǻ).

Figure 2.

The interactions between the lead ligands and the kinase domain of HER2. (A) – Eupatolitin, (B) - Rhamnetin, (C) - Annulatin, (D) - Laricitrin, (E) – Sterubin

Meanwhile, Annulatin formed 3 hydrogen bond interactions with the kinase domain of the target receptor. Specifically, hydrogen bond donor contacts were made with Asn850 (1.78 Ǻ) via the hydroxyl group on position 7, Asp863 (1.98 Ǻ) via the hydroxyl group attached to the carbon on position 5′, and Ala751 (2.25 Ǻ) also via a hydroxyl group attached to the carbon on position 3′(Figure 2C).Laricitrin interacted with the target receptor with the same combination of interactions observed in Eupatolitin and Rhamnetin (Figure 2D). A hydrogen bond donor (2.31 Ǻ) via the hydroxyl group on position 3 and a hydrogen bond acceptor (2.05Ǻ) via the carbonyl group on position 4. Also, hydrogen bond donating contacts were made with Asp808 via the hydroxyl group on position 4’ (1.64 Ǻ) and position 5’ (1.65 Ǻ). Sterubin exhibited a slightly different trend of interactions. A hydrogen bond acceptor contact (1.89 Ǻ) formed via the carbonyl group on position 4 (Figure 2E). Additionally, Sterubin formed two hydrogen bond contacts with Asp808 via the hydroxy group on position 4 (1.58 Ǻ) and on position 5 (1.62 Ǻ).

3.3. Prime MM/GBSA

The docked HER2-ligand complexes were subjected to an MM/GBSA calculation for rescoring. MM/GBSA approach was used to calculate the binding energy, which is a computational approximation of the degree of spontaneity of HER2-O-methylated flavonoids interactions.²⁰

The result of the computational calculations showed that Eupatolitin, the top-scoring compound, exhibited the highest binding energy of the five O-methylated flavonoids with a ΔG_bind value of -56.918 kcal/mol (Figure 1B). However, Neratinib, an FDA-approved drug, is predicted to be the most spontaneous of all compounds tested, with a binding energy of -69.814 kcal/mol. Interestingly, Rhamnetin, which comes second in the molecular docking ranking, exhibited a binding energy of -52.249 kcal/mol. Further, Myricetin derivatives, Annulatin and Laricitrin, returned ΔG_bind values of -47.794 kcal/mol and -48.297 kcal/mol, respectively. Meanwhile, all the test compounds returned higher binding energies than Afatinib (-41.351 kcal/mol), another FDA-approved HER2 inhibitor. According to this in-silico calculation, the predicted spontaneity of the eight interactions tested (five O-methylated flavonoids and three FDA-approved HER2 inhibitors) is predicted to be in the order Neratinib > Eupatolitin > Mubritinib > Rhamnetin > Laricitrin > Annulatin > Sterubin > Afatinib (Figure 1B). Molecular dynamics simulation, a more sophisticated approach, was employed to study the stability of the complexes formed from the interactions between the top-scoring O-methylated flavonoids and the kinase domain of HER2.

3.4. Molecular Dynamics Simulation

3.4.1. Ligand-HER2 Complex Stability and Amino Acid Residue Stability Analysis for 100ns

Figure 3 summarizes the molecular dynamics behavior of the five HER2-ligand complexes over 100 ns, including RMSD, RMSF, radius of gyration, hydrogen-bond profiles, and SASA.

Figure 3.

(A) Root mean square deviation of the docked complexes, (B) The root mean square fluctuation (RMSF) value of the hit compounds, (C) Radius of gyration, (D) Hydrogen bond analysis, (E) HER2 solvent accessible surface area

The RMSD profiles showed that all five HER2-ligand complexes remained stable over the 100 ns simulation period, although they differed in the extent and timing of their fluctuations (Figure 3A). Annulatin showed the most stable trajectory after equilibration, with the smallest overall deviation. Rhamnetin and Sterubin also maintained comparatively stable profiles, with Sterubin reaching equilibrium early in the simulation. In contrast, Eupatolitin and Laricitrin displayed increased mobility at later stages of the trajectory, indicating greater conformational flexibility after initially stable behavior.

The RMSF results were consistent with the RMSD analysis (Figure 3B). Eupatolitin, Rhamnetin, and Laricitrin showed similar residue-level fluctuation patterns around the active-site region, whereas Sterubin and Annulatin induced larger local fluctuations. Overall, these findings suggest that all five ligands formed stable HER2 complexes, but with measurable differences in local flexibility and dynamic behavior

3.4.2. The Radius of Gyration Analysis and Hydrogen Bond Analysis

The Rg profiles indicated only modest differences in HER2 compactness among the five complexes (Figure 3C). Annulatin showed the most stable Rg pattern throughout the simulation, while Rhamnetin and Sterubin remained relatively steady with only minor fluctuations. Eupatolitin showed slight early compaction followed by stable behavior, whereas Laricitrin displayed a transient increase in flexibility during the middle of the simulation before returning to a more stable range. Overall, the Rg values suggest that ligand binding did not induce major structural disruption of HER2.

Hydrogen-bond analysis showed that Annulatin and Laricitrin reached the highest maximum hydrogen-bond counts, whereas Eupatolitin and Rhamnetin maintained more consistent hydrogen bonding over time (Figure 3D). Sterubin formed fewer hydrogen bonds overall. These results indicate that the compounds differ not only in the number of hydrogen bonds formed, but also in the temporal persistence of those interactions.

3.4.3. Protein Solvent Accessible Surface Area (SASA)

SASA analysis showed moderate differences in solvent exposure among the HER2-ligand complexes over the simulation period (Figure 3E). Eupatolitin showed an early decrease in SASA, consistent with a more compact protein conformation during the initial phase of the simulation. Rhamnetin remained comparatively stable throughout, while Annulatin exhibited a decrease followed by a later increase in SASA. Laricitrin and Sterubin showed similar early trends but diverged at later stages of the trajectory. Overall, the SASA profiles support the RMSD and Rg results by indicating that all five complexes remained structurally stable, with only moderate differences in compactness and solvent exposure.

3.5. Density Functional Theory and Theoretical Quantum Chemical Calculations

The electronic properties of the top-scoring O-methylated flavonoids were ascertained using a Density Functional Theory (DFT) method.³¹

The calculated electronic properties include the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO), otherwise known as the Frontier molecular orbitals (Figure 4, Supplemental Figure 1). Different molecular properties and global reactivity parameters were calculated from the optimized geometries in addition to the Molecular electrostatic potential (MEP) surface (Supplemental Figure 1). Molecular electrostatic potential is a plot of electrostatic potential over a constant electron density of molecules. Through a distinguishable color scheme, the molecular electrostatic potential provides a pictorial representation of the size, shape, and charge of molecules.³² The calculated properties include EHOMO, ELUMO, ΔEgap, ionization potential (I), electron affinity (A), chemical potential (µ), chemical hardness (η), softness (S), electronegativity (χ), and electrophilicity (ω) (Table 2). According to Koopman’s theorem, these parameters can be defined in terms of the energy of the frontier orbitals: HOMO and LUMO.³³ Laricitrin exhibited the highest HOMO energy (-5.40 eV). Sterubin returned the most negative EHOMO value (-5.83 eV). Sterubin returned the highest ELUMO value (-1.35 eV), while Eupatolitin returned an ELUMO value of -1.85 eV. The difference in energy between HOMO and LUMO is the energy gap.³⁴

Figure 4.

3D plots of the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) for the 5 hit compounds

Table 2.

DFT-Derived Frontier Orbital Energies (eV) and Global Reactivity Descriptors of the Top Five O-Methylated Flavonoids

Parameter	Derivation	Eupatolitin	Rhamnetin	Annulatin	Laricitrin	Sterubin
E_HOMO (eV)	$= E_{H O M O}$	–5.46	–5.42	–5.65	–5.40	–5.83
E_LUMO (eV)	$= E_{L U M O}$	–1.85	–1.80	–1.73	–1.80	–1.35
ΔE_gap (eV)	$E_{L U M O} - E_{H O M O}$	3.61	3.62	3.92	3.60	4.48
A (eV)	$(A) = - E_{L U M O}$	1.85	1.80	1.73	1.80	1.35
I (eV)	$(I) = - E_{H O M O}$	5.46	5.42	5.65	5.40	5.83
η (eV)	$(ŋ) = \frac{I - A}{2}$	1.81	1.81	1.96	1.80	2.24
S (eV)^-1	$(S) = \frac{1}{2 ŋ}$	0.28	0.28	0.26	0.28	0.22
χ (eV)	$(χ) = \frac{I + A}{2}$	3.66	3.61	3.69	3.60	3.59
ω (eV)	$(ω) = \frac{µ^{2}}{2 ŋ}$	3.70	3.60	3.47	3.60	2.88
µ (eV)	$(µ) = - χ$	–3.655	–3.61	–3.69	–3.60	–3.59

Electron affinity (A), Ionization potential (I), global hardness (ŋ), global softness (S), electronegativity (χ), electrophilicity (ω), Chemical potential (µ).

Ionization potential (I) and Electron affinity (A) can be defined based on the values of E_HOMO and E_LUMO.³³ Sterubin exhibited the highest ionization potential (5.83 eV). Laricitrin returned the lowest ionization potential (5.40 eV). Eupatolitin exhibited the highest electron affinity (1.85 eV). Sterubin returned the lowest electron affinity (1.35 eV) (Table 2). Electronegativity measures the degree to which an atom, group of atoms or molecule attracts electrons.³⁵ Annulatin has a higher electronegativity than the other O-methylated flavonoids tested. Sterubin exhibited the lowest electronegativity (3.59 eV). Annulatin showed the lowest Chemical potential (-3.69 eV). Sterubin returned a µ value of -3.59 eV. Electrophilicity (ω) predicts probability of charge donation. Eupatolitin exhibited the highest electrophilicity (3.70 eV). Rhamnetin and Laricitrin returned an electrophilicity value of 3.60 eV (Table 2).

3.6. Predicted Pharmacokinetic Profiles of Reported O-Methylated Flavonoids

A structure-based pharmacokinetic screening was carried out on the top-scoring compounds to analyze their drug-like behavior (Table 3, Supplemental Table 1). ESOL model of water solubility was employed to predict the solubility behavior of the test compounds. Laricitrin exhibited the highest level of water solubility with an ESOL Log S value of -3.21. Rhamnetin and Sterubin exhibited water solubility levels comparable to those of Laricitrin, with Log S values of -3.36 and -3.47, respectively. Annulatin and Eupatolitin had the lowest recorded water solubility with values of -3.61 and -3.81, respectively. Sterubin returned the highest lipophilicity with a Log P value of 1.86. Laricitrin returned the lowest lipophilicity (Log P = 1.3). Eupatolitin, Rhamnetin, and Annulatin exhibited lipophilicity values of 1.72, 1.63, and 1.35, respectively. As predicted by SwissADME, Lipinski, Ghose, Veber, Egan, and Muegge rule-based filters were employed to predict the drug-likeness of the lead compounds. Eupatolitin, Rhamnetin, and Sterubin obeyed all the rule-based filters of drug-likeness. Annulatin and Laricitrin had one violation each in the Veber and Egan section of drug-likeness: TPSA > 140 (140.59 Å²) and TPSA > 131.6 Å². All the compounds returned a Bioavailability score of 0.55.

Table 3.

Predicted Physicochemical Properties, Water Solubility, and Drug-likeness Descriptors of the Top Five O-Methylated Flavonoids

	Eupatolitin	Rhamnetin	Annulatin	Laricitrin	Sterubin
Physicochemical properties
Molecular weight (g/mol)	346.29	316.26	332.26	332.26	302.28
Consensus Log P	1.72	1.63	1.35	1.3	-3.47
ESOL Log S	-3.81	-3.36	-3.61	-3.21	1.86
ESOL class	Soluble	Soluble	Soluble	Soluble	Soluble
Drug-likeness & Bioavailability
Lipinski #violations	0	0	0	0	0
Ghose #violations	0	0	0	0	0
Veber #violations	0	0	1	1	0
Egan #violations	0	0	1	1	0
Muegge #violations	0	0	0	0	0
Bioavailability score	0.55	0.55	0.55	0.55	0.55

3.7. Position of O-Methylation Enhances HER2 Binding Compared With Parent Flavonoids

To understand the impact of methoxylation on the HER2 antagonizing property of flavonoids, the parent flavonoids of the 5 hit compounds were also docked against HER2. O-methylation of Quercetin at positions 6 and 7 to form Eupatolitin (6-methoxy-7-methylquercetin) increases the binding affinity score from -9.518 kcal/mol to -10.092 kcal/mol. Rhamnetin, another Quercetin derivative, exhibited a slight increase in binding affinity from -9.518 kcal/mol to -9.990 kcal/mol. (Table 4). Meanwhile, the binding affinity of Annulatin (a Myricetin molecule methoxylated at position 3) is lower than the parent flavonoid, Myricetin. Similarly, Laricitrin, another Myricetin derivative, returned a lower binding affinity (-9.569 kcal/mol) than Myricetin (-10.013 kcal/mol).

Table 4.

Comparison of Glide XP Docking Scores (kcal/mol) for Selected O-Methylated Flavonoids and Their Parent Flavonoid Backbones Against HER2

O-Methyl flavonoids	Docking score	Flavonoid backbone	Docking score
Eupatolitin	-10.092	Quercetin	-9.518
Rhamnetin	-9.990	Quercetin	-9.518
Annulatin	-9.572	Myricetin	-10.013
Laricitrin	-9.569	Myricetin	-10.013
Sterubin	-9.260	Eriodictyol	-8.967

To rationalize these differences, we studied the structures of the top 5 hit compounds to gain insight into how the structures of these compounds influence the binding to the HER2 protein. Rhamnetin and Sterubin each bear a 7-O-methyl substituent, whereas Eupatolitin carries 6-O- and 7-O-methyl groups (Figure 5A). By contrast, Laricitrin and Annulatin are substituted at 2-O- and 2′-O-positions, respectively. On this basis, we hypothesize that O-methylation on the A ring at C-6 and C-7 is associated with more favorable binding to the HER2 kinase domain, whereas O-methylation at C-2 or C-2′ is less influential. Consistent with this trend, Eupatolitin showed the most favorable docking score among the top 20 selected compounds.

Figure 5.

(A) Flavonoid backbone, (B) The binding energies of flavonoids and 6,7-O-methylated flavonoids

3.8. Comparative Binding Profiles of Four Flavonoids and Their 6,7-O-Methylated Derivatives

Based on the initial molecular docking results, we suggested that methoxylation at position 6 and 7 on ring A (Figure 5A) of flavonoids improved binding. Due to the findings, four flavonoids, namely Quercetin, Myricetin, Luteolin, and Eriodictyol, which constitute the backbone of the majority of the selected O-methylated flavonoids, were chosen for further analysis. The structures of these compounds with methoxy groups at positions 6 and 7 were modeled using Marvin Sketch, optimized, and docked against the kinase domain of HER2 to unravel the impact of 6,7–O-methylation on the binding affinity, binding energy, and drug-likeness.

Docking results indicate that O-methylation at C-6 and C-7 consistently yields more favorable (more negative) HER2 binding energies across all four scaffolds (Table 5). Quercetin improved from -9.518 to -10.092 kcal/mol. Myricetin improved from -10.013 to -10.364 kcal/mol. Luteolin improved from -9.100 to -9.479 kcal/mol. Eriodictyol improved from -8.972 to -9.243 kcal/mol. Overall, 6,7-O-methylation enhanced the predicted HER2 affinity by approximately 0.27 to 0.57 kcal/mol relative to the parent flavonoids.

Table 5.

Comparison of Glide XP Docking Scores (kcal/mol) for Selected Flavonoids and Their Corresponding 6,7-O-methylated Derivatives Against HER2

Flavonoids	Docking score	Docking score
Quercetin	-9.518	-10.092
Myricetin	-10.013	-10.364
Luteolin	-9.100	-9.479
Eriodictyol	-8.972	-9.243

3.9. Prime MM/GBSA

Consistent with the docking results, Prime MM/GBSA estimates indicated stronger HER2 binding after adding methoxy groups at C-6 and C-7 for all four scaffolds (Figure 5B). Quercetin improved from -46.945 to -56.918 kcal/mol (Δ = -9.973). Myricetin improved from -47.577 to -56.920 kcal/mol (Δ = -9.343). Luteolin improved from -45.349 to -51.554 kcal/mol (Δ = -6.205). Eriodictyol improved from -37.001 to -44.100 kcal/mol (Δ = -7.099). Overall, 6,7-methoxy substitution increased ΔG_bind by ∼6–10 kcal/mol, with more negative values indicating stronger predicted binding.

3.10. Effects of O-Methylation on Selected Structure-based Pharmacokinetic Descriptors

The effects 6,7-O-methylation on the ADMET behaviors of test flavonoids were also examined (Supplemental Table 2). The absorption descriptors were significantly affected by the 6,7-O-methylation of Quercetin, Myricetin, Luteolin, and Eriodictyol. Quercetin, upon methylation (Supplemental Table 3), returned an increase in Log P value from 1.23 to 1.72.

4. Discussion

In the search for new therapeutics for various diseases, attention is constantly drawn to natural compounds with the hope of identifying compounds with higher activity and very little or no side effects.³⁶ In this study, 68 naturally occurring O-methylated flavonoids were screened against the HER2 kinase domain to identify compounds with favorable predicted binding (Supplemental Table 4). About 45% of the compounds returned docking scores of ≤ -7.0 kcal/mol, and 34 compounds fell within the range of -7.0 to -10.10 kcal/mol (Figure 1A). The top 20 compounds are summarized in Table 1, and five were selected for further analysis. Eupatolitin showed the most favorable docking score of all screened compounds and outperformed the reference HER2 inhibitors used in this study. Rhamnetin, Annulatin, Laricitrin, and Sterubin also showed stronger docking scores than Neratinib, Afatinib, and Mubritinib.

The combined docking and MM/GBSA workflow supports comparative ranking of the screened ligands. MM/GBSA provides an estimate of binding free energy (ΔGbind), with more negative values indicating more favorable predicted binding. In this analysis, Eupatolitin showed the most favorable MM/GBSA value among the tested flavonoids, suggesting stronger predicted HER2 engagement than the other lead compounds. Rhamnetin also showed a favorable binding energy, and all five selected flavonoids outperformed Afatinib in the MM/GBSA comparison. These results support the docking-based prioritization of the lead compounds. MM/GBSA is an approximate post-docking rescoring method that depends on the quality of the input poses and typically uses simplified implicit-solvent models. In addition, entropic contributions, protein conformational sampling, and force-field dependence may influence calculated values. Accordingly, the MM/GBSA results reported here are most appropriately interpreted as relative indicators of binding favorability rather than definitive free-energy estimates. Therefore, these computational binding energies were not used to infer biological activity directly, but only to support relative ranking and prioritization of the flavonoids for future experimental testing.

Molecular dynamics simulations provided additional insight into the stability of the HER2-ligand complexes over 100 ns. Overall, all five top-scoring O-methylated flavonoids formed stable complexes with HER2, although they differed in the extent of their conformational fluctuations. Annulatin showed the most stable RMSD and Rg behavior, while Rhamnetin and Sterubin also remained comparatively stable after equilibration. Eupatolitin and Laricitrin exhibited greater mobility at later stages of the simulation, suggesting increased conformational flexibility rather than loss of binding. Although Eupatolitin exhibited greater RMSD fluctuations at later stages of the simulation, its relatively favorable MM/GBSA binding energy may reflect the persistence of energetically favorable intermolecular interactions over the sampled trajectory. These fluctuations do not necessarily indicate loss of binding affinity but may instead reflect local structural adjustment or flexible accommodation within the HER2 binding pocket. The hydrogen-bond and SASA results supported these patterns: Annulatin and Laricitrin reached the highest maximum hydrogen-bond counts, whereas Eupatolitin and Rhamnetin maintained more persistent interactions over time. Overall, these MD results support stable HER2 engagement by all five compounds.

Hydrogen-bond analysis showed that Annulatin and Laricitrin formed the highest maximum numbers of hydrogen bonds, suggesting comparatively extensive intermolecular interactions with the HER2 kinase domain. Eupatolitin and Rhamnetin had the same number of hydrogen bonds, which showed that their interactions were stable. Sterubin, on the other hand, had the fewest hydrogen bonds. SASA analysis showed that the surface area of all complexes changed only slightly. Eupatolitin showed a drop in SASA early on, which meant that its structure was getting more compact. Laricitrin and Sterubin showed similar trends early on, but they went in different directions later in the simulation. Rhamnetin stayed about the same, but Annulatin grew in size in the later stages. The changes in SASA match up well with the RMSD and Rg profiles, which backs up the differences in how stable and compact the different complexes are. Overall, the MD simulations support stable HER2 engagement by all top-scoring O-methylated flavonoids, with Annulatin and Rhamnetin showing the most stable dynamic profiles, while Eupatolitin maintained favorable interactions despite greater late-stage fluctuations. These results support the MMGBSA-based binding energy calculations and show that these compounds could be good HER2 kinase domain inhibitors.

The frontier molecular orbital and global reactivity data provide a useful electronic context for the MM/GBSA and molecular dynamics results. In general, a smaller HOMO-LUMO energy gap is associated with higher chemical reactivity, whereas a larger gap suggests greater intrinsic stability. These descriptors, therefore, help explain why some compounds show stronger predicted HER2 binding and more stable interaction patterns than others.

Laricitrin had the highest HOMO energy and the lowest ionization potential, indicating comparatively strong electron-donating ability. Its MD profile suggested stable binding during the early part of the simulation, followed by moderate later fluctuations. Eupatolitin showed high electron affinity and the highest electrophilicity, consistent with its favorable MM/GBSA value of -56.918 kcal/mol. Together with its mostly stable RMSD profile, this suggests strong HER2 binding with some late-stage local conformational adjustment.

Sterubin showed the highest ionization potential, the highest chemical hardness, and the lowest electron affinity, indicating lower reactivity in the unbound state. This is consistent with its moderate simulation stability and weaker MM/GBSA binding energy. Annulatin combined high electronegativity with low chemical potential and showed the most stable MD profile, supporting its ability to maintain stable binding with limited structural disturbance. Rhamnetin showed balanced HOMO-LUMO properties and a favorable MM/GBSA value, consistent with its overall stable interaction pattern.

The residue-level hydrogen bonding interactions from molecular docking make these patterns clearer. Eupatolitin interacted with Met801 via donor (2.25 Å) and acceptor (2.04 Å) contacts, and with Asp808 through two donor interactions (1.66 Å and 1.69 Å). The strong binding energy and stable MD profile could be a result of this large network of interactions. Rhamnetin showed a similar pattern, making donor interactions with Asp808 (1.63 Å and 1.60 Å) and both donor and acceptor contacts with Met801 (2.25 Å and 1.95 Å). Annulatin made hydrogen bonds with Asn850 (1.78 Å), Asp863 (1.98 Å), and Ala751 (2.25 Å), which involved a wider range of residues. This distributed bonding, and its low chemical potential, could explain why it had relatively strong structural stability during simulation. Laricitrin exhibited a comparable Met801/Asp808 interaction pattern to Eupatolitin and Rhamnetin (2.31 Å, 2.05 Å, 1.64 Å, 1.65 Å). Sterubin had a simpler interaction profile, with one acceptor interaction at Met801 (1.89 Å) and two donor interactions at Asp808 (1.58 Å and 1.62 Å). This more limited hydrogen-bonding pattern may be related to its less favorable MM/GBSA binding energy.

These results show how electronic structure (DFT descriptors), binding energy (MM/GBSA), dynamic behavior (MD simulation), and residue-level contacts are all integrated. Eupatolitin and Rhamnetin are examples of compounds that have strong electrophilicity and large networks of hydrogen bonds. These compounds are predicted to interact with HER2 more strongly and spontaneously. Annulatin made more hydrogen bonds at different residues and a high electronegativity, which keeps its structure very stable. On the other hand, Sterubin, which is very hard and binds relatively poorly, has a lower binding strength and less dynamic adaptation. This integrated analysis provides a basis for understanding structure-property relationships in the HER2-O-methylated flavonoid systems and computationally prioritizes Eupatolitin and Annulatin for further experimental evaluation.

The pharmacokinetic screening reveals the oral drug-likeness of the highest-scoring O-methylated flavonoids. Laricitrin had the best water solubility, and Eupatolitin had the lowest. Sterubin had the highest lipophilicity, which means it was more likely to cross the intestinal lumen. Eupatolitin, Rhamnetin, and Sterubin all followed the rules for drug-likeness.^37-41 However, Annulatin and Laricitrin each broke one of the rules in the Veber and Egan filters because their TPSA values were too high. All the compounds got a bioavailability score of 0.55, which means they have good pharmacokinetic potential.

The comparative binding analysis suggests that O-methylation at C-6 and C-7 enhances flavonoid binding to HER2, whereas natural O-methylation at C-2 or C-2′ does not produce the same effect. Dual methylation, as observed in Eupatolitin,⁴² produced the strongest predicted binding. This suggests that A-ring substitution may be important for maintaining favorable interactions within the HER2 kinase domain. Overall, strategic 6,7-O-methylation appears to be a useful approach for optimizing flavonoid-based HER2 ligands by improving predicted target engagement while retaining favorable drug-like properties.

Previous HER2-related flavonoid studies support the idea that flavonoid scaffolds can interfere with HER2 signaling and kinase function.^43,44 For example, luteolin has been reported to promote HER2 degradation and inhibit proliferation in HER2-overexpressing cancer cells, while quercetin has been shown to downregulate HER2/neu and suppress its tyrosine kinase activity. In addition, hesperetin and naringenin have been reported to act as HER2 tyrosine kinase inhibitors, and prior structure-based screening studies have identified natural compounds as plausible binders of the HER2 kinase domain.^45-48 Our findings are consistent with this broader literature but add a more specific structure-affinity interpretation: in the present dataset, favorable HER2 binding was most strongly associated with O-methylation at C-6 and C-7, whereas methylation at other positions was not consistently beneficial. This suggests that the substitution pattern, and not simply flavonoid class alone, may be an important determinant in the rational optimization of HER2-directed flavonoid scaffolds.

Although this study is entirely computational, it provides an integrated assessment of ligand binding, energetic favorability, molecular stability, and electronic properties in the HER2 binding site. Docking reliability was supported by co-crystallized ligand redocking and comparison with known HER2 inhibitors; however, additional validation using independent docking software or cross-docking would further strengthen future studies. Another limitation is that the MD simulations were performed as single 100 ns trajectories for each HER2-ligand complex because of computational resource constraints; therefore, the RMSD, RMSF, Rg, SASA, and hydrogen-bond results should be interpreted as trajectory-based indicators of complex stability rather than statistically replicated estimates. These findings strengthen the rationale for the proposed compounds and support their prioritization for further development. Nevertheless, computational evidence alone is not sufficient to establish biochemical HER2 inhibition, cellular activity, selectivity over related kinases, pharmacodynamic properties, or in vivo efficacy. In addition, the 6,7-O-methylated derivatives examined here remain predictive designs until they are synthesized and experimentally validated. Accordingly, these computational results should be examined experimentally, with in vitro HER2 kinase inhibition assays as the immediate next step, followed by validation in HER2-positive breast cancer cell models.

5. Conclusion

This study identifies C-6/C-7 O-methylation as the clearest structural feature associated with improved predicted HER2 binding among the flavonoids examined. Eupatolitin emerged as the strongest overall computational hit, while Rhamnetin, Annulatin, Laricitrin, and Sterubin also showed favorable binding profiles. Overall, the docking, MM/GBSA, and molecular dynamics results support position-specific O-methylation as a rational strategy for optimizing flavonoid-based HER2 ligands. These computational findings warrant experimental validation, particularly through in vitro HER2 kinase inhibition assays, followed by HER2-positive cell-based studies to confirm the predicted activities of the lead compounds.

Supplemental Material

Supplemental material - Integrated Computational Profiling Links Position-specific O-Methylation to Spontaneous Complexation and Improved Binding in Flavonoid–HER2 Systems

Supplemental material for Integrated Computational Profiling Links Position-specific O-Methylation to Spontaneous Complexation and Improved Binding in Flavonoid–HER2 Systems by Precious A. Akinnusi, Samuel O. Olubode, David N. A. Okai, Oluwatobi I. Akinbami, David O. Oyerinde, Ayomide J. Akinnusi in Bioinformatics and Biology Insights

Footnotes

Ethical Considerations

Not applicable. This study did not involve human participants, patient data, or animal experiments.

Authors’ Contribution

Conceptualization: P.A.A.

Methodology: P.A.A., S.O.O., D.O.O., D.N.O.,

Software: P.A.A., D.N.O., A.J.A.

Writing-original draft: P.A.A., O.I.A., A.J.A., Writing-review and editing: D.N.O., A.J.A., O.I.A, D.O.O.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data used to support the findings of this study are included within the article.*

Supplemental Material

Supplemental material for this article is available online.

References

Lukasiewicz

Czeczelewski

Forma

Baj

Sitarz

Stanislawek

. Breast Cancer-Epidemiology, Risk Factors, Classification, Prognostic Markers, and Current Treatment Strategies-An Updated Review. Cancers (Basel). 2021;13(17):4287. doi:10.3390/cancers13174287.

Ferlay

Ervik

Lam

. Global cancer observatory: cancer tomorrow. Lyon, France: International Agency for Research on Cancer; 2020.

Kurebayashi

. Biological and clinical significance of HER2 overexpression in breast cancer. Breast Cancer. 2001;8(1):45-51. doi:10.1007/BF02967477.

Shah

Osipo

. Cancer stem cells and HER2 positive breast cancer: The story so far. Genes & diseases. 2016;3(2):114-123.

Pernas

Tolaney

. HER2-positive breast cancer: new therapeutic frontiers and overcoming resistance. Ther Adv Med Oncol. 2019;11:1758835919833519. doi:10.1177/1758835919833519.

Marra

Chandarlapaty

Modi

. Management of patients with advanced-stage HER2-positive breast cancer: current evidence and future perspectives. Nature reviews Clinical oncology. 2024;21(3):185-202.

Agostinetto

Curigliano

Piccart

. Emerging treatments in HER2-positive advanced breast cancer: Keep raising the bar. Cell Reports Medicine. 2024;5(6):101575.

Saleh

Khoury

Khalife

, et al. Mechanisms of action and resistance to anti-HER2 antibody-drug conjugates in breast cancer. Cancer Drug Resistance. 2024;7:22.

Wang

. Bioactive flavonoids in medicinal plants: Structure, activity and biological fate. Asian J Pharm Sci. 2018;13(1):12-23. doi:10.1016/j.ajps.2017.08.004.

10.

Kondratyuk

Pezzuto

. Natural product polyphenols of relevance to human health. Pharm Biol. 2004;42(sup1):46-63. doi:10.3109/13880200490893519.

11.

Kim

Lee

, et al. Regiospecific methylation of naringenin to ponciretin by soybean O-methyltransferase expressed in Escherichia coli. J Biotechnol. 2005;119(2):155-162. doi:10.1016/j.jbiotec.2005.04.004.

12.

Koirala

Thuan

Ghimire

Thang

Sohng

. Methylation of flavonoids: Chemical structures, bioactivities, progress and perspectives for biotechnological production. Enzyme Microb Technol. 2016;86:103-116. doi:10.1016/j.enzmictec.2016.02.003.

13.

Murakami

Koshimizu

Ohigashi

, et al. Characteristic rat tissue accumulation of nobiletin, a chemopreventive polymethoxyflavonoid, in comparison with luteolin. Biofactors. 2002;16(3-4):73-82. doi:10.1002/biof.5520160303.

14.

Morley

Ferguson

Koropatnick

. Tangeretin and nobiletin induce G1 cell cycle arrest but not apoptosis in human breast and colon cancer cells. Cancer Lett. 2007;251(1):168-178. doi:10.1016/j.canlet.2006.11.016.

15.

Wen

Walle

. Methylated flavonoids have greatly improved intestinal absorption and metabolic stability. Drug Metab Dispos. 2006;34(10):1786-1792. doi:10.1124/dmd.106.011122.

16.

Walle

Wen

Walle

. Improving metabolic stability of cancer chemoprotective polyphenols. Expert Opin Drug Metab Toxicol. 2007;3(3):379-388. doi:10.1517/17425255.3.3.379.

17.

Barreiro

Kummerle

Fraga

. The methylation effect in medicinal chemistry. Chem Rev. 2011;111(9):5215-5246. doi:10.1021/cr200060g.

18.

Aertgeerts

Skene

Yano

, et al. Structural analysis of the mechanism of inhibition and allosteric activation of the kinase domain of HER2 protein. J Biol Chem. 2011;286(21):18756-18765. doi:10.1074/jbc.M110.206193.

19.

Kim

Thiessen

Bolton

, et al. PubChem Substance and Compound databases. Nucleic Acids Res. 2016;44(D1):D1202-D1213. doi:10.1093/nar/gkv951.

20.

Akinnusi

Olubode

Alade

, et al. A molecular modeling approach for structure-based virtual screening and identification of novel anti-hypercholesterolemic agents from Grape. Informatics in Medicine Unlocked. 2022;32:101065. doi:10.1016/j.imu.2022.101065.

21.

Friesner

Murphy

Repasky

, et al. Extra precision glide: docking and scoring incorporating a model of hydrophobic enclosure for protein-ligand complexes. J Med Chem. 2006;49(21):6177-6196. doi:10.1021/jm051256o.

22.

Ajiboye

Akinnusi

Fatoki

, et al. In silico assessment of Hibiscus sabdariffa as a possible therapeutic agent for breast cancer management. Informatics in Medicine Unlocked. 2023;41:101330.

23.

Ajiboye

Fatoki

Akinnusi

, et al. Molecular docking, MMGBSA, and ADMET studies of phytoconstituents of Ocimum gratissimum on multiple breast cancer targets. Natural Product Research. 2025;39(16):4669-4677.

24.

Hayes

Archontis

. MM-GB (PB) SA calculations of protein-ligand binding free energies. In: Molecular dynamics-studies of synthetic and biological macromolecules. Rijeka, Croatia: InTech / InTechOpen; 2012:171-190.

25.

Pandian

SRK

Pavadai

Vellaisamy

, et al. Formulation and evaluation of rutin-loaded solid lipid nanoparticles for the treatment of brain tumor. Naunyn Schmiedebergs Arch Pharmacol. 2021;394(4):735-749. doi:10.1007/s00210-020-02015-9.

26.

Abraham

Murtola

Schulz

, et al. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX. 2015;1:19-25.

27.

Turner

. Center for Coastal and Land-Margin Research, Oregon Graduate Institute of Science and Technology. Beaverton, OR: Center for Coastal and Land-Margin Research, Oregon Graduate Institute of Science and Technology; 2005:2. XMGRACE, Version 5.1. 19.

28.

Shao

Molnar

Jung

, et al. Advances in methods and algorithms in a modern quantum chemistry program package. Phys Chem Chem Phys. 2006;8(27):3172-3191. doi:10.1039/b517914a.

29.

Banerjee

Eckert

Schrey

Preissner

. ProTox-II: A webserver for the prediction of toxicity of chemicals. Nucleic Acids Res. 2018;46:W257-W263.

30.

Daina

Michielin

Zoete

. SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules. Sci Rep. 2017;7(1):42717. doi:10.1038/srep42717.

31.

Summerfield

. Simulation of 5-Fluorouracil Intercalated into Montmorillonite Using Spartan’14: Molecular Mechanics, PM3, and Hartree-Fock. Open Journal of Physical Chemistry. 2015;5(03):49-55.

32.

Faloye

Bekono

Fakola

, et al. Elucidating the Glucokinase Activating Potentials of Naturally Occurring Prenylated Flavonoids: An Explicit Computational Approach. Molecules. 2021;26(23):7211. doi:10.3390/molecules26237211.

33.

Geerlings

De Proft

Langenaeker

. Conceptual density functional theory. Chem Rev. 2003;103(5):1793-1873. doi:10.1021/cr990029p.

34.

Bendjeddou

Abbaz

Gouasmia

Villemin

. Quantum chemical studies on molecular structure and reactivity descriptors of some P-nitrophenyl tetrathiafulvalenes by density functional theory (DFT). Acta Chim Pharm Indica. 2016;6(2):32-44.

35.

Pitzer

. The nature of the chemical bond and the structure of molecules and crystals: an introduction to modern structural chemistry. Journal of the American Chemical Society. 1960;82(15):4121.

36.

Akinnusi

Olubode

Salaudeen

. Molecular binding studies of anthocyanins with multiple antiviral activities against SARS-CoV-2. Bull Natl Res Cent. 2022;46(1):102. doi:10.1186/s42269-022-00786-0.

37.

Lipinski

Lombardo

Dominy

Feeney

. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv Drug Deliv Rev. 2001;46(3):3-26. 00129-0.

38.

Ghose

Viswanadhan

Wendoloski

. A knowledge-based approach in designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A qualitative and quantitative characterization of known drug databases. J Comb Chem. 1999;1(1):55-68. doi:10.1021/cc9800071.

39.

Veber

Johnson

Cheng

Smith

Ward

Kopple

. Molecular properties that influence the oral bioavailability of drug candidates. Journal of Medicinal Chemistry. 2002;45(12):2615-2623. doi:10.1021/jm020017n.

40.

Egan

Merz

Jr Baldwin

. Prediction of drug absorption using multivariate statistics. J Med Chem. 2000;43(21):3867-3877. doi:10.1021/jm000292e.

41.

Muegge

Heald

Brittelli

. Simple selection criteria for drug-like chemical matter. J Med Chem. 2001;44(12):1841-1846. doi:10.1021/jm015507e.

42.

Wang

Liao

Liu

Shi

. Diversity and regioselectivity of O-methyltransferases catalyzing the formation of O-methylated flavonoids. Critical Reviews in Biotechnology. 2024;44(6):1203-1225.

43.

Hossam

Abdallah

El-Sebakhy

Alnajjar

Mohyeldin

. Discovery of natural scaffolds as HER2 inhibitors for breast cancer: virtual screening, molecular dynamics, and biological characterization with selectivity profiling. Scientific Reports. 2025;15(1):25883.

44.

Prajoko

Qhabibi

Gerardo

, et al. Revealing novel source of breast Cancer inhibitors from seagrass Enhalus acoroides: in Silico and in vitro studies. Molecules. 2024;29(5):1082.

45.

Chiang

C-T

Way

T-D

Lin

J-K

. Sensitizing HER2-overexpressing cancer cells to luteolin-induced apoptosis through suppressing p21WAF1/CIP1 expression with rapamycin. Molecular cancer therapeutics. 2007;6(7):2127-2138.

46.

Jeong

Kwon

Lee

. Quercetin‐induced ubiquitination and down‐regulation of Her‐2/neu. Journal of cellular biochemistry. 2008;105(2):585-595.

47.

Chandrika

Steephan

Kumar

Sabu

Haridas

. Hesperetin and naringenin sensitize HER2 positive cancer cells to death by serving as HER2 tyrosine kinase inhibitors. Life sciences. 2016;160:47-56.

48.

Rampogu

Son

Baek

, et al. Targeting natural compounds against HER2 kinase domain as potential anticancer drugs applying pharmacophore based molecular modelling approaches. Computational Biology and Chemistry. 2018;74:327-338.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

1.64 MB

0.00 MB