Ligand configurational entropy and protein binding

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Ligand binding to proteins involves a loss of configurational entropy, penalizing binding affinity.
• Current computational models for ligand design vary in how they address this entropic penalty.
• The physical assumptions behind these entropic penalty calculations require validation.

Purpose of the Study:

• To investigate the physical basis of configurational entropy loss upon small molecule binding.
• To analyze the association of amprenavir with HIV protease using computational methods.
• To evaluate the accuracy of existing energy models in accounting for entropic penalties.

Main Methods:

• Utilized Mining Minima calculations to analyze ligand-protein association.
• Quantified the loss in ligand configurational entropy.
• Examined the contribution of energy well depth versus rotamer number to entropy loss.

Main Results:

• A significant loss in ligand configurational entropy was computed for amprenavir binding to HIV protease, contributing substantially to the overall binding free energy (DeltaG degrees).
• The primary driver of this entropy loss was identified as narrower energy wells in the bound state, rather than a reduction in accessible ligand rotamers.
• Coupling between rotational, translational, and internal degrees of freedom complicated the decomposition of entropy changes.

Conclusions:

• Conformationally restricted ligands offer a potential strategy for enhancing binding affinity.
• The findings have significant implications for refining energy models used in ligand scoring for drug design.
• Accurate modeling of entropic penalties is crucial for effective computer-aided drug design.