Identification of Cryptic Binding Sites Using MixMD with Standard and Accelerated Molecular Dynamics

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Protein dynamics are crucial for small molecule binding and identifying potential binding sites.
• Cryptic binding sites require significant protein structural rearrangement to become accessible to ligands.

Purpose of the Study:

• To evaluate the Mixed-solvent Molecular Dynamics (MixMD) protocol for identifying cryptic binding sites in proteins with diverse conformational changes.
• To assess the performance of MixMD in mapping sites requiring side-chain, loop, or larger structural movements.

Main Methods:

• Applied the MixMD protocol using molecular dynamics (MD) simulations of unbound proteins in a solution containing 5% probe molecules and explicit water.
• Utilized standard and accelerated MD simulations to enhance sampling of torsional angles.
• Performed docking studies using ensembles of protein conformations generated by MixMD.

Main Results:

• MixMD successfully mapped cryptic binding sites in 5 out of 12 proteins using standard simulations.
• Accelerated MD was required for partial mapping of cryptic sites in guanylate kinase and TIE-2.
• MixMD, even with accelerated dynamics, could not map sites requiring helix or domain movements within 100 ns timescales.
• Conformational dynamics in MixMD simulations were similar to water-only simulations, suggesting probes map inherent protein dynamics.
• Docking to standard MixMD ensembles outperformed apo crystal structures in 9 cases and half of bound structures.

Conclusions:

• MixMD is a valuable tool for identifying cryptic binding sites, particularly those involving moderate conformational changes.
• The probes in MixMD likely exploit spontaneous protein opening events rather than inducing them.
• Ensemble docking based on MixMD simulations offers improved accuracy over static structure docking for drug discovery.