Reconciling membrane protein simulations with experimental DEER spectroscopy data.

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Membrane protein structure determination is challenging, often relying on spectroscopy and simulations.
• Double electron-electron resonance (DEER) provides distance distributions, while molecular dynamics (MD) simulations reveal conformational dynamics.
• Existing methods struggle to reconcile DEER data with MD simulations, hindering validation and experimental design.

Purpose of the Study:

• To investigate the causes of discrepancies between DEER experiments and MD simulations for membrane proteins.
• To identify how membrane environment and spin label modification affect residue-pair distance distributions.
• To evaluate strategies for improving the agreement between experimental and simulation data.

Main Methods:

• Atomistic molecular dynamics (MD) simulations of membrane proteins (PepT_So and LeuT).
• Comparison of residue-pair distance distributions from simulations with experimental DEER data.
• Analysis of the impact of different membrane mimetics and covalent spin label modifications.

Main Results:

• The choice of membrane mimetic (detergent micelle) had minimal impact on protein dynamics.
• Covalent modification of residues to nitroxide spin labels significantly altered local dynamics and distance measurements.
• Measuring distances between spin-labeled oxygen atoms, rather than protein backbone, caused major divergence.

Conclusions:

• Covalent modification of residues for DEER experiments is the primary source of mismatch with MD simulations.
• Biased simulations focusing only on spin labels can distort underlying protein dynamics.
• This work identifies key factors causing DEER-MD discrepancies, enabling improved validation and experimental design for membrane proteins.