ProGuide: a flexible framework for modeling global conformational rearrangements in proteins using DEER-derived...

Area of Science:

• Structural Biology
• Biophysics
• Computational Chemistry

Background:

• Protein conformational heterogeneity is critical for function, necessitating experimental methods to visualize distinct states.
• Double electron-electron resonance (DEER) spectroscopy provides long-range distance information between spin labels, resolving protein conformations.
• Computational modeling of protein backbone rearrangements using DEER distance restraints is challenging due to spin label rotameric flexibility.

Purpose of the Study:

• To introduce ProGuide, a novel computational framework for generating accurate protein structural models guided by DEER distance distribution data.
• To address the challenge of modeling protein backbone rearrangements and spin label rotameric flexibility in DEER-guided modeling.
• To apply ProGuide to a G protein-coupled receptor (GPCR) to reveal distinct conformational states.

Main Methods:

• ProGuide utilizes iterative, experimentally biased molecular dynamics simulations to capture large conformational rearrangements.
• Spin-label rotameric heterogeneity is modeled using chiLife, and Cα changes are calculated to match experimental DEER distributions.
• A selection process generates an ensemble of models that best recapitulates the DEER data.

Main Results:

• ProGuide successfully generated accurate structural models for the angiotensin II type 1 receptor (AT1R) using published DEER data.
• The models revealed distinct Gq- and β-arrestin-biased conformations, including a novel β-arrestin-biased state.
• Structural insights into tertiary rearrangements and residue-level changes in microswitch motifs were obtained.

Conclusions:

• ProGuide effectively models protein conformational rearrangements using DEER-derived distance restraints.
• The framework demonstrates power and flexibility in investigating large, complex proteins.
• This approach advances the ability to generate complete structural models of multiple protein conformations.