Protein-Backbone Thermodynamics across the Membrane Interface.

Area of Science:

• Biophysics
• Computational Chemistry
• Membrane Protein Studies

Background:

• Protein-membrane interactions are crucial for biological processes.
• Accurate thermodynamic models are needed for transmembrane proteins.
• Coarse-grained (CG) models offer computational efficiency but require careful parametrization.

Purpose of the Study:

• To investigate the impact of backbone glycyl unit representations on protein insertion thermodynamics.
• To develop and validate a method for estimating atomistic potential of mean force (PMF) curves from CG simulations.
• To compare the performance of Martini and PLUM CG models against atomistic simulations and experimental data.

Main Methods:

• Utilized computer simulations to model protein insertion into lipid membranes.
• Developed a novel methodology to derive atomistic PMF curves from CG snapshots.
• Performed atomistic simulations for reference data and validation.
• Compared two prominent CG models: Martini and PLUM.

Main Results:

• Identified significant discrepancies between Martini and PLUM models compared to atomistic PMFs and experiments.
• Atomistic simulations indicated a weak free energy of insertion for the glycyl unit into POPC membranes.
• Refining the backbone representation in PLUM substantially improved the accuracy of WALP16 peptide insertion PMFs.
• Demonstrated that imbalances in backbone-side chain contributions cause energetic artifacts in CG models.

Conclusions:

• Accurate parametrization of CG models requires careful consideration of backbone contributions.
• The PLUM model, with refined backbone parametrization, shows improved accuracy for membrane insertion thermodynamics.
• Free-energy-based parametrization of single-residue models presents challenges due to the complex nature of partitioning free energies.