(203h) G?-Models for RNA Folding

In the last decade, the folding of biomolecules has been studied using long timescale molecular dynamics (MD) simulations on special purpose supercomputers. While these simulations have provided invaluable insights about biomolecular folding, even high-performance supercomputing machines are limited to simulating the folding of relatively small biomolecules. To study such rare events, various computational methods for accelerating and enhancing sampling during MD simulations have been designed. Here, we used structure-based Gō-model simulations to characterize the folding of a transfer RNA (tRNA) molecule with and without the N²,N²-dimethylguanosine (m^2,2G) modification. This modification is known to prevent misfolding of tRNA which can otherwise lead to a decrease in the translation efficiency and to the development of various diseases. We observed that the folding of the modified tRNA proceeded in a cooperative hierarchal manner with a specific order of folding among various structural motifs. However, in the absence of m^2,2G we observed distinct folding pathways of tRNA that may result in misfolded configurations of tRNA. Therefore, m^2,2G likely restricts the folding landscape of tRNA directing it to a more favorable folding pathway. These results highlight the significance of conformational effects in tRNA folding originating from chemical modifications to its nucleotides.