Simulations of Trp-cage miniprotein show cryo-EM vitrification shifts conformer populations only slightly

Researchers have used molecular dynamics to test how the rapid cooling in cryo-EM sample preparation changes the distribution of protein conformations. The work, published in Proceedings of the National Academy of Sciences, covers more than 50 milliseconds of simulations of the Trp-cage miniprotein, both at equilibrium and during cooling doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗. In single-particle cryo-EM the sample is plunged into a cryogen and vitrified, which takes about 22 microseconds. Until now the effect of this step on the ensemble of biomolecules was unknown doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗.

Simulated cooling rates and water behavior

The authors simulated seven cooling rates spanning three orders of magnitude. The slowest of them match experimental rates that still result in vitrification doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗. They examined molecular mobility and density-temperature equations of state for water with and without the protein. From these they concluded that the vitrification of water is not altered by the protein doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗.

To track conformational changes and relate them to kinetics, the team built a Markov state model (MSM) of Trp-cage. It was based on 5.4 milliseconds of equilibrium sampling at 277 K doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗.

Population shifts and recovery of equilibrium

MSM states whose characteristic time exceeds the duration of the nonequilibrium cooling were generally more robust to cooling artefacts doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗. In the equilibrium ensemble at 230 K some states vanish. None vanished in the ensembles produced by nonequilibrium cooling doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗. According to the authors, the population shifts caused by cooling are small but observable doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗.

For more labile states, which are perturbed more strongly by cooling, the authors developed a thermodynamic inference framework. It recovers equilibrium populations from the measured vitrified ensembles doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗. On this basis they conclude that cryo-EM can serve as a reliable and accurate biophysical technique for studying biomolecular ensembles doi.orgProceedings of the National Academy of Sciencesdoi.orgOpen source ↗. The simulations were carried out for a single miniprotein, Trp-cage.