Insights into nucleic acid conformational dynamics from massively parallel stochastic simulations

Area of Science:

• Biophysics
• Computational Biology
• Molecular Biology

Background:

• Helical hairpins are fundamental nucleic acid secondary structures with crucial functional roles.
• The atomic-level self-assembly dynamics of these structures remain incompletely understood.

Purpose of the Study:

• To investigate the dynamics of nucleic acid hairpin formation and disruption at the atomic level.
• To characterize rare events in hairpin folding previously unobservable in simulations.

Main Methods:

• Employed large-scale, parallel, atomistic molecular dynamics simulations on a distributed computing system (>40,000 processors).
• Collected over 500 microseconds of simulation time for a hairpin ensemble (sequence 5'-GGGC[GCAA]GCCU-3').
• Utilized uncoupled ensemble dynamics, conformer-specific folding probabilities, and multiplexed replica exchange stochastic dynamics.

Main Results:

• Identified competing pathways between folded and unfolded conformational states.
• Observed nonnative stacking and basepairing interactions acting as folding traps.
• Characterized a distinct helix unwinding-rewinding dynamic mode separate from standard folding/unfolding.
• Derived an approximate folding landscape revealing a heterogeneous transition state ensemble.

Conclusions:

• Nucleic acid hairpin folding involves complex dynamics, including competing pathways and nonnative traps.
• A novel helix unwinding-rewinding mechanism contributes to the overall folding process.
• While native topology influences the folding landscape, polymer chemistry dictates pathway probabilities, contrasting with peptide beta-hairpins.