Protein-induced membrane curvature in coarse-grained simulations

Area of Science:

• Biophysics
• Computational Biology
• Membrane Biology

Background:

• The endosomal sorting complex required for transport (ESCRT)-III complex is crucial for membrane remodeling.
• Peripheral proteins play key roles in membrane curvature sensing and generation.
• Accurate computational models are needed to understand these protein-membrane interactions.

Purpose of the Study:

• To evaluate three coarse-grained MARTINI models (regular, POL-MARTINI, BMW-MARTINI) for simulating ESCRT-III's membrane remodeling activities.
• To compare the models' ability to capture protein-protein and protein-membrane interactions.
• To assess their performance in predicting membrane curvature sensing and generation.

Main Methods:

• Utilized three variants of the MARTINI coarse-grained model: regular, POL-MARTINI, and BMW-MARTINI.
• Simulated a linear filament model of ESCRT-III.
• Compared simulation results with all-atom and atomistic simulations.

Main Results:

• All three MARTINI models consistently described the ESCRT-III protein-protein interface.
• Significant differences were observed in protein-membrane interactions and membrane curvature behaviors across models.
• BMW-MARTINI showed the best agreement with all-atom simulations for the protein-membrane interface.
• Regular MARTINI qualitatively matched atomistic simulations for membrane curvature sensing and generation.
• POL-MARTINI exhibited weak protein-membrane interactions, lacking curvature-sensitive activities.
• BMW-MARTINI's inaccuracies stemmed from overestimated amphipathic helix insertion and incorrect spontaneous curvature of anionic lipids.

Conclusions:

• Coarse-grained models offer different strengths and weaknesses in simulating membrane remodeling.
• BMW-MARTINI and regular MARTINI provide valuable but distinct insights into ESCRT-III function.
• Model improvements are needed, focusing on lipid composition and protein motif insertion depth for accurate curvature regulation.