Innovations In in Silico Assessment of Quercetin: Anti - Inflammatory Perspectives

Increased levels of cytokines, cytokine receptors, adhesion molecules, and various immunoregulatory mediators are the hallmarks of inflammation, a complex host (systemic or local) reaction to tissue damage and infection. By controlling the release of pro-inflammatory cytokines like IL - 1α, IL - 1β, and IL - 6 as well as chemokines like IL - 8 and RANTES, histamine - a biogenic amine primarily released from mast cells and basophils - plays a crucial role in inflammatory responses [1], [2]. Histamine has a major role in both acute hypersensitivity responses and allergic and inflammatory reactions.  It functions through several histamine receptors (H1-H4), which regulate activity of immune cells and the production of cytokines during inflammation [3], [4].  Additionally, by causing immune and endothelial cells to generate pro-inflammatory cytokines like IL-6 and IL-8, histamine can heighten inflammatory reactions [5].  Although inflammation is an essential defense mechanism, excessive or chronic inflammation is associated with a number of disorders, include arthritis, cardiovascular issues, diabetes and neurological diseases. Finding effective and safe anti-inflammatory medications has therefore become more crucial in biomedical research.

Numerous conditions, including the rheumatoid arthritis, asthma, IBD and hypersensitivity, have been linked to inflammation. Due to serious side effects and their unsuitability for long-term use, some marketed therapeutic medications have been discontinued. Consequently, managing inflammatory diseases remains a major challenge for medical professionals [6].

Flavonoids, which are naturally occurring polyphenolic compounds present in fruits, vegetables, and medicinal plants, are well known for their strong anti - inflammatory and antioxidant properties [1], [7]. Among them, quercetin is widely distributed in food sources and possesses a wide range of pharmacological activities. Its ability to suppress the activation of transcription factors like nuclear factor-κB (NF-κB), inhibit the production of pro-inflammatory cytokines like TNF-α, interleukin-6 (IL-6), and interleukin-1β (IL-1β), and modulate inflammatory signalling pathways are the main reasons for its anti-inflammatory effects [8], [9.

The development of computational biology has made in silico techniques useful for drug development and discovery. Researchers can forecast how bioactive compounds will interact with particular target proteins involved in inflammatory pathways through in silico analysis, which includes molecular docking, molecular dynamics simulations, and pharmacokinetic predictions. Before performing laboratory experiments, these computational methods assist in assessing the binding affinity, stability, and possible inhibitory effects of compounds, saving time and resources [10].

Thus, the current study aims to explore the anti-inflammatory property of quercetin through in silico approach by assessing its interaction with some major anti-inflammatory target proteins. This might give an idea about the mechanism of action of quercetin, thereby making it an anti-inflammatory therapeutic agent.

MATERIALS AND METHODS

Molecular docking

Molecular docking was done to analyse the binding energies of Quercetin with P38 MAPK, IL4, IL6, TNFα, IL 1β, PDE4D, TLR 4, ERK 2, and NF - KB. Ligand structure was downloaded from PubChem in .sdf format and then converted into .pdbqt file by using Open babel: (The Open-Source Chemistry Toolbox). Receptor preparation was done using UCSF Chimerax 1.9, CHARMM-GUI, Autodock and AGFR. Docking was done by using Autodock 4.2.6. The docked structures were visualized by using Discovery Studio Visualizer v24.1.0.23298.

MD simulation

MD simulations were carried out with the GROMACS package, and the CHARMM36 force field was applied. Energy minimization was performed with the steepest descent method, and a threshold of 1000 kJ mol?¹ nm?¹ was used to remove steric clashes. Equilibration was performed in two steps. First, the system was equilibrated in the NVT ensemble at a temperature of 300 K using the V-rescale thermostat with a τ of 0.1 ps. Next, the system was equilibrated in the NPT ensemble at a pressure of 1 bar with the Parrinello - Rahman barostat, where τ was 2.0 ps and compressibility was 4.5 × 10?? bar?¹.

For electrostatics, the Particle Mesh Ewald (PME) algorithm was used with a grid spacing of 0.16 nm and fourth - order interpolation. For van der Waals interaction calculations, a cutoff of 1.2 nm was used along with a force switch modifier. For hydrogen bonds, constraints were applied using the LINCS algorithm with fourth - order expansion of the constraint matrix. Simulations were run for 100 ns with a time step of 2 fs and trajectory information saved every 10 ps. Periodic boundary conditions were used to ensure system continuity, and the Verlet cutoff scheme was used to optimize neighbour searching.

The analyses performed after the simulations include the following: RMSD, RMSF, RG, hydrogen bonds, PCA Analysis, and the calculation of the binding free energies by the MM/PBSA method.

RESULTS AND DISCUSSION

Molecular docking analysis

The molecular docking analysis showed that the binding affinities of quercetin with different target proteins vary, and the range of the binding energy is between -3.94 and -6.13 kcal/mol. Among the inflammatory mediators, the highest binding affinity was observed with p38 MAPK (-6.13 kcal/mol), followed by IL-1β (-6.02 kcal/mol), PDE4D (-5.99 kcal/mol), TNFα (-5.88 kcal/mol), TLR4 (-5.85 kcal/mol), and IL-4 (-5.08 kcal/mol). A low level of binding energy is indicative of stronger binding. Further, the visualization of the docked complex revealed that the p38 MAPK established four hydrogen bonds (Table 1), as well as hydrophobic interactions. These hydrogen bonds and hydrophobic interactions enabled the stable position of the quercetin complex.

Table 1: Quercetin in silico docking findings with several inflammatory mediator proteins

Fig 1A- 1F: Important hydrogen bonding and nonbonding interactions in the docked complex between Quercetin and ligand

The molecular docking technique was used to determine the potential of quercetin in its role as a multi - target inhibitor of inflammatory mediators. This was performed using the AutoDock GPU package, and the data showed a range of binding energies from -2.99 to -6.13 kcal/mol for the nine target proteins. This indicated a favourable interaction with the protein targets, with the most favourable interaction being with p38 mitogen-activated protein kinase (-6.13 kcal/mol, PDB ID: 3GCU). This was followed by a very favourable interaction with interleukin-1β (-6.02 kcal/mol, PDB ID: 5I1B), phosphodiesterase 4D (-5.99 kcal/mol, PDB ID: 5K32), tumor necrosis factor-α (-5.88 kcal/mol, PDB ID: 2AZ5), and Toll-like receptor 4 (-5.85 kcal/mol, PDB ID: 2Z62). Weaker interactions were also noted with interleukin-4 (-5.08 kcal/mol, PDB ID: 2NVH), extracellular signal-regulated kinase 2 (-3.94 kcal/mol, PDB ID: 2OJI), interleukin-6 (-3.60 kcal/mol, PDB ID: 1N26), and nuclear factor - κB (-2.99 kcal/mol, PDB ID: 1NFK).

The number of hydrogen bonds varied between 2 and 5. PDE4D and TNFα showed the maximum number of 5 Hydrogen bonds. This could be due to the presence of various hydroxyl groups on the quercetin ring, which could be interacting with the polar residues like aspartates/glutamates present in the active sites. Visualization of the docked compounds (Fig 1A - 1F) revealed interesting details. For the p38 MAPK enzyme (3GCU), four hydrogen bonds were accompanied by hydrophobic interactions of the drug with residues like Leu108 and Tyr35. The drug was docked into the ATP binding site. Likewise, in the case of the IL-1β (5I1B), four bonds were formed that included residues such as Gln15 and Ser43, plays an important function in responses to inflammation as well as plays a role in the activation of numerous signalling pathways related with immune regulation [11]. Hence, Quercetin's interaction with IL-1β may help decrease inflammatory signals. PDE4D (5K32) also showed five bonds that included residues of the catalytic site of the enzyme, such as His160 and Asp275. Phosphodiesterase-4 enzymes catalyze the hydrolysis of the cyclic adenosine monophosphate (cAMP). Inhibition of PDE4D limits cAMP breakdown, which consequently decreases inflammatory mediator production and minimizes inflammatory reactions [12]. This could be the cause of the inhibition of the hydrolysis of cAMP and the resulting inflammation. These results are consistent with the previous reports [8].

In the case of the TNFα (2AZ5), five bonds were formed that included residues Tyr59 and Gln61. TNF-α is a trimeric cytokine that interacts with TNF receptors. Inhibiting its trimer formation can limit downstream inflammatory signalling pathways [13]. Quercetin's interaction with these residues could interfere with trimer stability and suppress TNF α - mediated inflammation. Similarly, TLR4 (2Z62) showed four bonds that included with Arg264 and Glu439, hindering the recognition of lipopolysaccharide. TLR4 is a pattern recognition receptor that detects lipopolysaccharide (LPS) in Gram-negative bacteria and triggers innate immune responses. Quercetin binding to these residues may impair LPS recognition, hence inhibiting TLR 4 activation and lowering pro - inflammatory cytokine production [14].

While, IL 4 (2NVH) showed weaker affinity with two hydrogen bonds. The lower connection between quercetin and IL-4, which regulates anti-inflammatory and immunological responses, indicating that it may have a limited direct influence on this cytokine compared to stronger inflammatory mediators like TNF α and TLR4. By altering important cytokines and immunological signalling pathways, our results validate quercetin's unique anti-inflammatory potential [8].

Because of their higher binding energies (< -5.85 kcal/mol) and the amount of hydrogen bonds (> 4), which suggested inhibitory potential, 3GCU, 5I1B, 5K32, 2AZ5, and 2Z62 were chosen for MD simulations.

MD stimulation analysis

100 ns MD simulations were done using the GROMACS software and the force field to evaluate the docked complexes' dynamical behaviour and long-range stability. The simulation protocol includes energy minimization using the steepest descent method (convergence criterion of 1000 kJ/mol/nm), NVT ensemble simulations at 300 K with the V-rescale thermostat (τ=0.1 ps), and NPT ensemble simulations at 1 bar using the Parrinello-Rahman barostat (τ=2.0 ps and compressibility=4.5×10?? bar?¹). Electrostatics were treated with Particle Mesh Ewald (PME, grid spacing = 0.16 nm), van der Waals with a cutoff at 1.2 nm and force switch, and hydrogen bonds were constrained with LINCS, fourth - order expansion. Simulations were run with a 2-fs time step, and configurations were saved every 10 ps under periodic boundary conditions and with Verlet neighbor search.  The structural stabilities were also evaluated using root mean square deviation (RMSD), root mean square fluctuation (RMSF), radius of gyration (Rg), and solvent accessible surface area (SASA).

The RMSD curves (Fig 3) showed that the simulations equilibrated quickly within the range of 20-50 ns. The RMSD values were generally <0.3 nm for all the complexes, suggesting that no major conformational shifts. PDE4D-quercetin (5K32) had the lowest RMSD range, averaging between 0.15 and 0.2 nm, suggesting robust docking predictions and little drift throughout simulations.  Strong intermolecular contacts and favourable ligand lodging within the binding pocket are widely thought to be demonstrated by this stability [15].  The existence of flexible sections inside the hinges of kinases may be the cause of the somewhat greater the RMSD spectrum for p38 MAPK-quercetin (3GCU), which is around 0.25 nm on average [16].  Reduced atomic fluctuations inside bind pockets (0.1–0.2 nm) were seen in the RMSF analysis (Fig. 4), with PDE4D-quercetin exhibiting the greatest stiffness increase, perhaps as a result of π-stacking among quercetin as well as its aromatic residues.  It has been demonstrated that these interactions enhance ligand stability and significantly increase flavonoids' propensity for binding to enzyme sites linked to inflammatory processes [17].

Rg trajectories (Fig 4) were consistent with structural compactness, with values converging to 1.8–2.2 nm across all systems. This implies that no unfolding/expansion occurs. Stable Rg values are frequently viewed as markers of the protein backbone's the integrity of structures and compactness during simulated trajectories [15]. The system with the most consistent Rg value is the TNFα-quercetin complex (2AZ5) ~2.0 nm. This implies that the hydrophobic core is preserved. The minor variations found in the Rg profiles show that the protein maintained its compact structure after ligand interaction. This stability implies that quercetin binding does not lead to significant structural changes to the protein structure. Moreover, in molecular dynamics simulations, the retention of hydrophobic packing inside the protein interior frequently leads to stable binding of ligands and overall complex stability [18].

The solvent accessible surface area (SASA) profiles (Fig 4) were analysed to evaluate the changes in protein surface exposure throughout the molecular dynamic’s simulation of ligand-protein complexes. The results revealed that a 10 – 15 % reduction in SASA upon binding, especially for the IL 1β - quercetin complex (5I1B), corresponding to a 120 - 140 nm² reduction. This implies that the solvent is excluded from the hydrophobic interfaces, increasing the entropic favourability [19].

These findings support the complexes' dynamical stability, with PDE4D - quercetin emerging as the stiff, consistent with its anti-inflammatory properties.

Analysis of binding free energy

The binding free energies were computed using the MM/PBSA and MM/GBSA methods for the five protein-ligand complexes. The MM/PBSA binding free energies varied between -28.33 ± 15.40 kJ/mol (2AZ5) and -96.65 ± 30.33 kJ/mol (5K32), confirming the strongest binding of PDE4D-quercetin, driven by the dominant ΔVdW (van der Waals) values of -96.27 ± 20.71 kJ/mol and ΔEele (electrostatic) values of -211.50 ± 66.40 kJ/mol. These were compensated by the unfavourable and large polar solvation penalty ΔGPB values of 224.81 ± 30.00 kJ/mol. These contributions are compatible with the structural properties of Quercetin with numerous hydroxyl groups that assist hydrogen bonding as well as electrostatic interactions, as well as a planar aromatic scaffold that facilitates hydrophobic stacking interactions [20].

The binding of IL-1β-quercetin (5I1B), -51.59 ± 13.39 kJ/mol, was balanced between ΔVdW and ΔEele. This balanced energetic pattern is indicative of ligands that attain optimum complementarity inside the binding pocket, hence increasing binding stability without incurring excessive energetic penalties [21]. The binding of TLR4-quercetin (2Z62), -38.87 ± 12.76 kJ/mol, and p38 MAPK-quercetin (3GCU), -42.09 ± 20.79 kJ/mol, was of medium strength.  These intermediate values indicate that quercetin is partially stabilized inside the active sites, most likely due to poor interaction networks or lower hydrophobic packing when compared to PDE4D. Nonetheless, these interactions remain physiologically significant, as both TLR4 and p38 MAPK are important mediators of pathways of inflammatory signalling, and even mild suppression may lead to anti-inflammatory effects [23], [24]. The weakest binding was that of TNFα - quercetin (2AZ5), possibly due to higher solvation costs (ΔGPB: 156.36 ± 19.71 kJ/mol) (Table 2). The lower affinity is mostly due to a substantial polar solvation penalty (ΔGPB = 156.36 ± 19.71 kJ/mol), that offsets beneficial intermolecular interactions. Quercetin's significant desolvation cost indicates that it might not be able to effectively replace water molecules or reach sufficient immobilization below the  binding site, which would lower its binding efficacy.  Polar ligands incur large energy penalties during the shift between solvent to protein environment, a feature that is commonly seen in MM/PBSA research [21].

Because of lower polar solvation penalties (ranging from 72.97 kJ/mol to 206.19 kJ/mol) and non-polar terms (ranging from -9.83 kJ/mol to -18.79 kJ/mol) that more closely approximate buried surface areas, the MM/GBSA results had a greater exothermic as well, with values for standard atoms between -45.06 ± 13.14 kJ/mol over 2Z62 to -120.37 ± 24.27 kJ/mol for 5K32. While considerable difference in findings highlights PBSA's dependency on dielectric constants vs GBSA's efficiency, consistency in ranking between techniques, from lowest to maximum binding energy, shows PDE4D as the major target (Table 3). Large standard deviations highlight the significance of more sampling, particularly with regard to kinase (3GCU) along with receptor (2Z62) systems.  Quercetin's enhanced binding to PDE4D suggests that it may play a role in altering the cyclic AMP-dependent pathways of cellular inflammation, which are essential for regulating the immune system [12]. MD energy values have been shown to be less favourable than docking for some ligands; for example, 2AZ5 displayed a ΔG of -28.33 kcal/mol by MM/PBSA and -5.88 kcal/mol by the method of docking.

Table 2: MMPBSA (Molecular Mechanics–Poisson–Boltzmann Surface Area) of Quercetin

Table 3: MMGBSA (Molecular Mechanics–Generalized Born Surface Area) of Quercetin

 Analysis Amino Acid Decomposition Analysis

Critical residues were identified by analyzing the binding energy elements corresponding to ΔG_bind using per-residue energy decomposition (Figs. 2A–2J). Phe372 created a van der Waals interaction by π-π stacking to explain the large ΔEele for PDE4D-quercetin (5K32), while His160 and Asp275 were predicted to make a contribution <-10 kJ/mol through electrostatics. For polyphenolic compounds with planar ring structures, such as quercetin, those aromatic stacking interactions are well recognized for their function in ligand stabilization inside hydrophobic spaces [20].  These findings are in line with previous research that highlights the importance of crucial active-site residues in affecting ligand binding with a balance of polar regions and non-polar relations [21].  Hydrophobic groups (F126 & L52) were significant donors (~ -5 kJ/mol respectively) for TLR4-quercetin (2Z62), making up for its low ΔEele.  These findings demonstrate that quercetin stabilization in the TLR4 binding site is mostly caused by hydrophobic contacts rather than significant electrostatic interactions. Because TLR4 is essential for both initiating innate immune system responses and controlling inflammatory signalling pathways, even little binding to TLR4 has physiological significance [23].  Polar residues helped bind IL-1β-quercetin (5I1B) through Arg4 as well as Glu105.  While the negatively charged Glu105 may form complimentary charge-dipole & hydrogen bonding connections with the ligand-specific hydroxyl groups, the positively charged Arg4 is expected to significantly increase electrostatic contact with electron-rich regions of quercetin.

The IL-1β-quercetin combination contributes equally to electrostatic and van der Waals energy, according to the MM/PBSA results. Quercetin can create a network of stabilizing polar interactions at the binding site due to the presence of both positive and negative charged residues, which enhances binding selectivity. These interactions are characteristic of flavonoid-protein complexes, in which polar amino acids and hydroxyl-rich ligands engage in many hydrogen bonds [20].  Tyr59, which is probably associated with π-π or π-alkyl bonds involving quercetin's planar ring structure, which helps contribute its van der Waals stability, and Gln61, the polar residue, may promote the hydrogen bonding and dipole-dipole reactions via quercetin's hydroxyl groups, thereby contributing an electrostatic component to the binding. In contrast, Tyr59 and Gln61 made balanced contributions for TNFα-quercetin (2AZ5).

Since TNF-α is a crucial cytokine in proinflammatory signaling, even small interactions can have prolonged anti-inflammatory effects when combined with greater inhibition of other targets, such as PDE4D and IL 1β [14].  Despite having lower individual energies, Met109 & Lys53 contributed to the binding of p38 MAPK-quercetin (3GCU).  However, as is common in kinase-ligand interactions, the presence of both polar (Lys53) and hydrophobic (Met109) residues points to a mixed interaction profile.  p38 MAPK is a key regulator of inflammatory signalling pathways, namely the production of pro-inflammatory cytokines [22].  Relatively mild binding interactions can therefore aid in the suppression and subsequent modulation of inflammatory processes.  Overall, our study verifies quercetin's activity through crucial residue interaction and pinpoints hot spots for mutagenesis.

2A: MMPBSA of 5K32_Quercetin

2B: MMGBSA of 5K32_Quercetin

2C: MMPBSA of 2Z62_Quercetin

2D: MMGBSA of 2Z62_Quercetin

2E: MMGBSA of 5I1B_Quercetin

2F: MMGBSA of 5I1B_Quercetin

2G: MMGBSA of 2AZ5_Quercetin

2H: MMGBSA of 2AZ5_Quercetin

2I: MMGBSA of 3GCU_Quercetin

2J: MMGBSA of 3GCU_Quercetin

Fig 2A - 2J: Amino acid analysis of Quercetin with targets

Hydrogen bonding analysis

Surprisingly, however, MD Hydrogen Bonding Analysis (Figs. 3 and 4) revealed "No H Bonds" for every complex, indicating the absence of stable and enduring hydrogen bonds throughout the simulations.  This stands in stark contrast to the docking data, which expected four to five bonds, such as the PDE4D complex's five bonds.  These differences are somewhat uncommon as MD simulations show the dynamic character of a system in a solvated and thermally changing environment, whereas docking provides a static image under idealized circumstances [16].  Strong ΔEele values, like -211.50 kJ/mol within the 5K32 complex, may, however, indicate alternative polar interactions that support the complexes, like salt bridges and/or water-mediated interactions, which are often missed because of the direct protein-ligand hydrogen bonds in conventional investigations [21].  Quercetin's dependence on hydrophobic as well as van der Waals forces under physiological conditions-possibly at the price of specificity however in accordance with broad-spectrum activity-may also be indicated by the absence of hydrogen bonds.  This is typical of polyphenolic substances, which favor dispersion interactions over rigid, directed hydrogen bonds due to the planar aromatic structures [20].  As a result, during MD simulations, the hydrogen bonding research demonstrates a shift from the static, hydrogen bond-like interactions expected in docking to a greater dynamic, hydrophobically driven binding mode.  This highlights how important it is to integrate many computational approaches in order to have a practical understanding of ligand-protein interactions.

Fig 3: Result of RMSD ANALYSIS

No H Bonds

Fig 4: Result of RMSF/RG/SASA/H BOND

Principal component Analysis (PCA) and Free Energy Landscape (FEL)

Apo proteins displayed a wider diversity of conformational configurations (variance > 30 %) than bound states, according to the analysis of principal components (PCA) 2D projections (Fig. 5A-5J).  With variance above 30 %, the apo proteins displayed a broader variety of conformations, suggesting more intrinsic flexibility and structural diversity in the lack of ligand interaction. In contrast, quercetin-bound complexes exhibited considerably limited conformational sampling, indicating the ligand binding stabilizes certain conformational states while reducing total protein mobility.

PDE4D - quercetin had the tightest clusters (variance < 15%), suggesting conformational locking. This tight clustering clearly supports a significant conformational limitation upon ligand binding, implying a "conformational locking" effect. Such behaviour is consistent with a very stable protein-ligand complex in which quercetin binding effectively constrains the protein's large-scale collective motion. Conformational locking is frequently associated with increased inhibitory potency, as the ligand keeps the protein in a certain conformational state that may be detrimental to its biological activity. In the case of PDE4D, such limitation of dynamic mobility could cause problems with substrate binding and catalytic activity, which contributes to its anti-inflammatory actions through modification of cAMP signalling pathways [13].

On the other hand, p38 MAPK - quercetin had a broader distribution, suggesting flexibility. The free energy landscape (Fig 6A - 6J) revealed that bound states had global minima. This shows the quercetin binding does not completely limit the dynamic movements of p38 MAPK, most likely due to poorer or less optimal interaction networks, as seen by its relatively low binding free energy and reduced residue-level contributions.  Because conformational plasticity is often retained to fulfil regulatory and catalytic tasks, kinase systems are renowned for their adaptability [22].

PDE4D - quercetin and IL 1β - quercetin had deeper minima (ΔG ~ -10 to -15 kJ/mol), with high energy barriers (> 5 kJ/mol), suggesting a stably bound state.  The presence of excellent as well as persistent thermodynamic conformational states is suggested by these enormous energy pools.  Furthermore, transitions beyond these states are energetically unfavourable due to the presence of rather large barriers to energy (> 5 kJ/mol) close to these minima, which increases the persistence of the bound conformations during the simulation [16].

However, TLR4-quercetin and TNFα-quercetin indicate transitory binding, which is associated with larger RMSD and lower ΔG_bind.  Reduced electrostatic contributions and insufficient residue-level stability are examples of poor interaction networks that might be responsible for the transient nature of binding in these systems.  Although Phe126 & Leu52 (TLR4) and Tyr59 & Gln61 (TNF-α) help in binding, their energy contributions are insufficient to anchor quercetin in a specific conformation.  Moreover, the total interaction is weakened by the absence of long-lasting hydrogen bonds and higher solvation penalties.  This behaviour is in line with the concept of "promiscuous binding," which is seen in many natural objects and where a small amount of affinity across several locations may have a significant impact on intricate biological processes [24].

5A: 5K32	5B: 5K32_Quercetin
5C: 2Z62	5D: 2Z62_Quercetin
5E: 5I1B	5F: 5I1B_Quercetin
5G: 2AZ5	5H: 2AZ5_Quercetin
5I: 3GCU	5J: 3GCU_Quercetin

Fig 5A - 5J: Result of PCA 2D Projections

6A: 5K32	6B: 5K32_Quercetin
6C: 2Z62	6D: 2Z62_Quercetin
6E: 5I1B	6F: 5I1B_Quercetin
6G: 2AZ5	6H: 2AZ5_Quercetin
6I: 3GCU	6J: 3GCU_Quercetin

Fig 6A – 6J: Result of Free Energy Landscape

Dynamic Cross-Correlation Matrix Analysis

Residue motion correlations were shown using dynamic cross-correlation matrices (DCCM) (Figs. 7A–7E).  While ligand-bound proteins, particularly the PDE4D-quercetin and IL 1β-quercetin complexes, showed higher beneficial associations (0.6-0.8) in the active site areas, suggesting dynamic stabilization, apo proteins showed sporadic positive/negative correlations (±0.4-0.6).  Quercetin's allosteric modulation is demonstrated by decreased anticorrelations seen in ligand-bound states, which may attenuate inflammatory signalling cascades.  The increase in correlated motion suggests that quercetin helps stabilize the internal dynamics of these proteins, promoting regulated fluctuations that are often associated with conformations that are both structurally and functionally stable.

Furthermore, inside the ligand-bound states, anticorrelated motions dramatically diminished. Often linked to structural alterations necessary for biological activity, anticorrelations are opposing movements between distant regions of a protein.  Their decline with quercetin binding suggests a reduction in extensive structural changes, which may restrict the flexibility required for protein function.  This pattern demonstrates feedback regulation, in which the structure & activity of distant regions are impacted by ligand binding at one location [25].

Fig 7A – 7E: DCCM analysis