Comparative binding energy analysis considering multiple receptors: a step toward 3D-QSAR models for multiple...

Area of Science:

• Computational Chemistry
• Medicinal Chemistry
• Drug Discovery

Background:

• Quantitative Structure-Activity Relationship (QSAR) models typically focus on a single receptor.
• Predicting both drug affinity and selectivity across related targets remains a challenge.
• Existing methods struggle to account for subtle differences in binding sites among homologous proteins.

Purpose of the Study:

• To extend comparative binding energy analysis for deriving multi-target 3D-QSAR models.
• To develop a unified model that simultaneously predicts affinity and selectivity for structurally related receptors.
• To investigate the key interactions driving selectivity for thrombin, trypsin, and factor Xa.

Main Methods:

• Incorporation of multiple structurally related receptors into the QSAR model's X-matrix.
• Application of the extended technique to a series of 88 3-amidinophenylalanines.
• Analysis of binding to thrombin, trypsin, and factor Xa (fXa) using 202 protein-ligand complexes.

Main Results:

• A single predictive regression model was generated for all three receptors with a Q(2) of 0.689.
• Ligand occupancy of key binding pockets (D-site, P-site, S1-site rim) correlates with affinity.
• Thrombin's unique 60-loop enhances P-site interactions, conferring specificity; D-site occupancy favors fXa inhibition; negative charge density near position 88 is crucial for thrombin recognition.

Conclusions:

• The extended 3D-QSAR approach successfully integrates affinity and selectivity prediction for multiple targets.
• Specific binding site features, like thrombin's 60-loop and Lys insertion, dictate selectivity.
• The model provides valuable insights for designing selective inhibitors targeting thrombin, trypsin, and fXa.