Are current atomistic force fields accurate enough to study proteins in crowded environments?

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Cellular interiors are crowded with macromolecules, affecting protein thermodynamics and kinetics.
• High protein concentrations are used in structural biology methods like NMR and spectroscopy.
• Classical molecular dynamics (MD) simulations' accuracy at high protein concentrations is largely untested.

Purpose of the Study:

• To investigate protein behavior and aggregation at high concentrations using MD simulations.
• To assess the accuracy of current MD force fields in mimicking crowded cellular environments.
• To study wild-type and oxidatively damaged villin headpiece aggregation.

Main Methods:

• Explicit-solvent MD simulations totaling 6.4 µs.
• Simulations of villin headpiece at 6 mM and 9.2 mM protein concentration.
• Utilized GROMOS 45a3 and 54a7 force fields with varying electrostatics and ionic strengths.

Main Results:

• Both wild-type and damaged villin headpiece showed similar aggregation, contrary to differing propensities.
• Wild-type villin headpiece aggregated even when experimentally soluble, across multiple simulation protocols.
• Aggregation correlated with a significant decrease in potential energy, involving hydrophobic, polar, and backbone interactions.
• This aggregation artifact was observed across multiple major atomistic force fields (AMBER, CHARMM, OPLS).

Conclusions:

• Current MD force fields may overestimate protein-protein interactions, leading to inaccurate aggregation predictions in crowded environments.
• MD simulations might distort the understanding of protein behavior in biologically relevant crowded conditions.
• Further development of force fields is needed to accurately model protein behavior at high concentrations.