High-resolution reversible folding of hyperstable RNA tetraloops using molecular dynamics simulations

Area of Science:

• Computational Biology
• Biophysics
• Structural Biology

Background:

• Hyperstable RNA tetraloops are crucial for RNA structure and function.
• Noncanonical interactions stabilize these RNA motifs.
• Accurate prediction of RNA tertiary structure remains challenging.

Purpose of the Study:

• To develop and validate accurate molecular dynamics (MD) simulation methods for predicting RNA tetraloop structures.
• To assess the ability of novel RNA force field parameters to capture key structural features of hyperstable tetraloops.

Main Methods:

• De novo folding simulations of three hyperstable RNA tetraloops (UUCG, GCAA, CUUG) using replica exchange molecular dynamics.
• Initialization of simulations from unfolded states with explicit solvent and ions.
• Development and application of unique RNA parameters calibrated from experimental thermodynamic and kinetic data.

Main Results:

• Achieved angstrom-level root-mean-square deviation (rmsd) accuracy for de novo folded RNA tetraloop structures compared to experimental data.
• Successfully reproduced noncanonical loop-stabilizing interactions characteristic of hyperstable tetraloops.
• Demonstrated accurate capture of RNA flexibility, base stacking energetics, and purine syn-anti interconversions.

Conclusions:

• The developed RNA parameters and MD simulation approach enable accurate prediction of RNA tertiary structure.
• This represents a significant advancement for computational prediction of RNA folding and function.
• Unbiased all-atom molecular dynamics simulations can now reliably recapitulate key RNA structural motifs.