In Silico Screening and Molecular Dynamics Validation of Bryophyllum Pinnatum Flavonoids as Pi3kγ Inhibitors For Rheumatoid Arthritis

Technical Summary: In Silico Screening and Molecular Dynamics Validation of Bryophyllum Pinnatum Flavonoids as PI3Kγ Inhibitors for Rheumatoid Arthritis

Problem Statement
Rheumatoid arthritis (RA) is a chronic autoimmune disorder characterized by progressive joint destruction, driven in part by the aberrant activation of the phosphatidylinositol 3-kinase gamma (PI3Kγ) isoform in immune cells and synovial fibroblasts. PI3Kγ promotes the infiltration of inflammatory cells (neutrophils and macrophages) and exacerbates the PI3K/AKT/mTOR pathway, leading to increased pro-inflammatory cytokines (TNF-α, IL-1β, IL-6) and synovial fibroblast overgrowth. While existing therapies exist, there is a need for novel, plant-based therapeutic agents. Bryophyllum pinnatum, a medicinal plant widely used in African ethnomedicine for inflammatory conditions, contains bioactive flavonoids, yet its specific molecular mechanism against PI3Kγ remains unexplored through computational validation.

Methodology
The study employed a comprehensive in silico workflow to evaluate flavonoids from B. pinnatum as potential PI3Kγ inhibitors (PDB ID: 2A4Z). The methodology proceeded through the following stages:

1. Protein and Ligand Preparation: The PI3Kγ structure was pre-processed (bond orders, hydrogen addition, loop filling) using the OPLS4 force field. A library of B. pinnatum compounds was retrieved from PubChem, converted to 3D, and ionized at pH 7.0 ± 0.2 using LigPrep.
2. Pharmacophore Modeling and Screening: An E-pharmacophore model was generated based on the co-crystallized ligand complex. This model, featuring seven hydrophobic groups and ten aromatic rings, was used to filter the compound library, retaining only those with at least two critical features.
3. Molecular Docking: The filtered compounds (25 total) were docked into the active site using the Glide Extra Precision (XP) algorithm. The binding site was defined by a grid box centered on the co-crystallized ligand.
4. ADMET Profiling: The top hits were evaluated for Absorption, Distribution, Metabolism, Excretion, and Toxicity (ADME) using the SwissADME server, specifically checking adherence to Lipinski's Rule of Five, polar surface area (TPSA), and skin permeability (logKp).
5. Binding Free Energy Calculations: Prime MM-GBSA was utilized to calculate the binding free energies (ΔGbind) of the top complexes to validate docking scores.
6. QSAR Modeling: An AutoQSAR model was developed using compounds from the ChEMBL database to predict pIC₅₀ values for the lead compounds.
7. Molecular Dynamics (MD) Simulation: A 100-nanosecond MD simulation was performed using the Desmond software (Schrödinger suite) with the OPLS4 force field and TIP3P solvent model. The stability of the protein-ligand complexes was assessed via Root Mean Square Deviation (RMSD) and Root Mean Square Fluctuation (RMSF), alongside interaction monitoring (hydrogen bonds, hydrophobic, ionic, and water bridges).

Key Results

• Docking Scores and Interactions: Three flavonoids—Luteolin, Quercetin, and Kaempferol—demonstrated superior inhibitory potential compared to the co-crystallized ligand (-9.628 kcal/mol) and the standard drug Tofacitinib (-6.205 kcal/mol).
  • Luteolin: -11.044 kcal/mol (Hydrogen bonds with LYS 833, VAL 882, GLU 880).
  • Quercetin: -10.723 kcal/mol (Hydrogen bonds with SER 806, VAL 882, LYS 833).
  • Kaempferol: -10.553 kcal/mol (Hydrogen bonds with SER 806, LYS 833, VAL 882; PI-PI stacking with TYR 867).
• ADMET Properties: All lead compounds satisfied Lipinski's Rule of Five (zero violations), exhibited favorable molecular weights, and possessed TPSA values below 140 Ų, indicating good oral bioavailability.
• Binding Free Energy (MM-GBSA): The lead compounds showed significantly stronger binding affinities than Tofacitinib.
  • Luteolin: -45.07 kcal/mol
  • Kaempferol: -44.79 kcal/mol
  • Quercetin: -44.03 kcal/mol
  • (Comparison: Co-crystallized ligand -44.57 kcal/mol; Tofacitinib -21.04 kcal/mol).
• QSAR Prediction: The AutoQSAR model (R² = 0.6646, Q² = 0.6625) predicted high inhibitory potency, with Luteolin showing the highest pIC₅₀ value (7.053), followed by Quercetin (6.952) and Kaempferol (6.627).
• MD Simulation Stability: Over 100 ns, the protein backbone RMSD remained generally below 3.6 Å, indicating structural integrity.
  • The Luteolin complex showed initial conformational adjustments before stabilizing.
  • The Quercetin complex stabilized after 20 ns.
  • The Kaempferol complex stabilized between 80–100 ns.
  • Tofacitinib required approximately 40 ns to stabilize.
  • Throughout the trajectory, the complexes maintained key interactions, including hydrogen bonds, hydrophobic contacts, and water bridges with residues such as LYS 833, VAL 882, and GLU 880.

Significance and Claims
The authors claim that this study provides a robust computational basis for the traditional use of Bryophyllum pinnatum in treating inflammatory conditions. The findings suggest that Luteolin, Quercetin, and Kaempferol are potent, stable, and drug-like inhibitors of PI3Kγ, potentially outperforming the standard drug Tofacitinib in terms of binding affinity and interaction stability.

The paper posits that these compounds could effectively interfere with PI3Kγ-mediated signaling pathways responsible for neutrophil/macrophage infiltration, cytokine overproduction, and synovial damage in RA. The study concludes that these in silico results offer a strong scientific rationale for the anti-RA properties of B. pinnatum constituents. However, the authors explicitly state that these findings are preliminary and require validation through in vitro kinase assays, cell-based models of synovial inflammation, and animal models of collagen-induced arthritis before clinical application.