Mining protein dynamics from sets of crystal structures using "consensus structures".

Area of Science:

• Structural biology
• Computational chemistry
• Drug discovery

Background:

• Large crystal structure datasets contain implicit dynamic information.
• Understanding protein dynamics is crucial for drug mechanism elucidation.
• Current methods often rely on single static structures.

Purpose of the Study:

• To develop novel methods for analyzing dynamic structural information from crystal data.
• To investigate ligand-induced changes and common binding features in HIV-1 reverse transcriptase.
• To identify new drug targets within the NNRTI binding pocket.

Main Methods:

• Visualizing ligand-induced changes using B-factors and residue displacements from multiple crystal structures.
• Generating "consensus structures" by overlapping multiple crystal structures to map common binding features.
• Analyzing protein-ligand interactions, hydrogen bonding, and binding pocket characteristics.

Main Results:

• Corroborated the dual mechanism of non-nucleoside reverse transcriptase inhibitors (NNRTIs) involving protein dynamics distortion and conformational changes.
• Identified a common wild-type NNRTI binding pocket.
• Discovered a conserved backbone hydrogen bond acceptor at P236 and a novel hydrophobic subpocket.

Conclusions:

• The novel methods effectively utilize data variability from multiple crystal structures.
• These approaches provide complementary insights beyond classical single-structure analysis.
• Findings offer potential for designing next-generation NNRTIs targeting conserved features.