Folding studies of hyperstable RNA tetraloops through molecular dynamics

Abstract

Within all living cells, processes of gene replication, metabolite sensing, and transcription/translation hinge on the folding of RNA molecules, from small stem-loops to large ribosomal fragments. The physico-chemical environment in which the RNA exists plays important roles in the processes of folding, and has been a subject of study for the biophysical community for many years.

The direct associations of RNA with its chemical environment are responsible for stabilizing multiple alternative configurations including left-handed helices in high concentrations of neutral salt, and unfolded configurations in denaturing cosolutes such as urea. Using concentrations of KCl (100 mM, 1 M, 2 M, 3 M) and high concentrations of urea (2 M, 4 M, 6 M), the preferential associations between these solution components with each other and the gcGCAAgc molecule are described in detail with isothermal (NVT) MD simulations. The ion associations match with experiments for similar concentrations and show saturation in the highest salt concentrations measured in our work. The associations of urea show that preferential associations between RNA and these denaturants match with observations of nucleobase-specific favorability from experiment, and additional analyses into the equilibrium stability of the model RNA stem-loop system show that urea preferentially disfavors formation of alternative configurations and reduces the variability of the configurational landscape.

These methodologies applied in these studies can describe the effects of solution environment on RNA and other biomolecules, and highlight the importance of assessing the folding of biomolecules as an ensemble of states rather than a single native folded state.

We characterize the heterogeneity of folding in a stem-loop-forming RNA molecule (gcGCAAgc) using enhanced sampling through replica-exchange molecular dynamics (REMD), and describe the configurational landscape with multiple structural metrics. The configurational landscape of the stem-loop in our studies is described by a complex ensemble of native and non-native folded states with differential stability relative to system temperature and pressure. Our observations provide insight into important properties of stem-loop folding including different folding rates based on different RNA interactions, ensembles of non-native stacking interactions in low temperatures, and interactions between RNA with ionic cosolutes.

Studies of folding for small RNA molecules with MD has typically focused on the formation of native configurations described by experiments like X-Ray crystallography and nuclear magnetic resonance spectroscopy (NMR) without considering the plethora of misfolded and non-native configurations suggested from experiments like temperature-jump and gel electrophoresis. Even the smallest stem-loop configurations have been described by multi-state ensembles, and the populations of different folded states are known to be affected by changes in temperature, pressure, ionic strength, and cosolute concentrations.

Increases in computational capacity have driven advances in computational-based analysis of RNA molecules, with a goal to using advanced molecular dynamics (MD) techniques to describe the folding and conformational stability of RNA molecules as a function of changes to the solution environment, and elucidating how these catalytic, information-storage molecules perform so many necessary functions in our cells. With recent modifications to existing MD force fields for RNA, it has become possible to recapitulate the folding of native RNA tetraloop configurations from unfolded precursors, but one of the long-standing, unanswered questions in MD analysis of RNA folding is describing the configurational variability and non-native ensembles for a single RNA molecule.