Ligand binding remodels protein side-chain conformational heterogeneity

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Protein conformational heterogeneity is crucial for biological functions like ligand binding but remains challenging to quantify.
• Macromolecular X-ray diffraction data, typically analyzed as static structures, can yield insights into both harmonic and anharmonic contributions to conformational dynamics.

Purpose of the Study:

• To quantitatively measure protein conformational heterogeneity in unbound versus ligand-bound states using multiconformer modeling.
• To investigate the relationship between ligand properties and protein flexibility changes upon binding.
• To explore how conformational heterogeneity propagates from the binding site in specific protein targets like CDK2.

Main Methods:

• Applied multiconformer modeling to analyze time- and space-averaged electron density from 743 matched crystallographic datasets (apo and holo states).
• Compared side-chain conformational heterogeneity between unbound and ligand-bound protein structures.
• Correlated observed protein flexibility changes with specific ligand properties (hydrogen bonds, hydrophobicity) and inhibitor features in CDK2 structures.

Main Results:

• Ligand binding induced increased flexibility in residues distant from the binding site when local binding site residues became more rigid, particularly in buried regions.
• Protein flexibility increased as the number of hydrogen bonds decreased and relative hydrophobicity increased with ligand properties.
• Conformational heterogeneity in CDK2 structures correlated with inhibitor features, demonstrating propagation of flexibility changes away from the binding site.

Conclusions:

• Findings support models where residual side-chain entropy modulates ligand binding affinity.
• Highlight the necessity of integrating both static conformational changes and dynamic conformational heterogeneity into ligand-binding models.
• Suggests X-ray diffraction, when analyzed with multiconformer modeling, is a powerful tool for characterizing protein dynamics relevant to function.