Excessive aggregation of membrane proteins in the Martini model

Area of Science:

• Computational biology
• Biophysics
• Molecular dynamics

Background:

• The Martini model is widely used for coarse-grained simulations of membrane proteins.
• Computational efficiency makes it suitable for studying protein oligomerization.
• However, it often predicts excessive and irreversible protein aggregation.

Purpose of the Study:

• To investigate the accuracy of the Martini model in predicting membrane protein oligomerization.
• To address the issue of exaggerated protein-protein interactions in the Martini model.
• To improve the reliability of Martini simulations for multi-protein complexes.

Main Methods:

• Comparison of dimerization free energies from Martini simulations with FRET experimental data.
• Assessment of Martini-predicted transmembrane domain dimer structures against NMR data.
• Modification of Martini protein-protein interaction parameters.

Main Results:

• Martini simulations showed excessive protein aggregation compared to FRET data.
• Martini-predicted dimer structures poorly matched NMR-resolved structures.
• Scaling down protein-protein interactions reduced aggregation but did not improve structural accuracy.

Conclusions:

• The Martini model exaggerates protein-protein interactions, leading to unrealistic aggregation.
• Adjusting interaction strength improves simulation realism but not structural prediction accuracy.
• Amino-acid specific interaction tuning is likely necessary for accurate structural predictions.