by Ansuman Lahiri | University of Calcutta
Post-transcriptional modifications occur extensively in RNA and provide an additional layer
of chemical coding that presumably alters the chemistry, structure and dynamics of the
molecules. Molecular dynamics simulations offer a convenient tool to elucidate the
mechanisms underlying the roles played by such modifications. However, since the
reliability of the simulation results depend on the accuracy of the force field employed, we
tested some of the available force field parameters for modified RNA residues for the
AMBER molecular modeling suite of programs. The results from our work revealed that the
available parameters from the AMBER distribution were inadequate in reproducing the
experimentally observed conformational distributions
We earlier adopted a procedure for reoptimizing the dihedral parameters of a few modified
uridines using high-level ab initio calculations. We earlier adopted a procedure for
reoptimizing the dihedral parameters of a few modified uridines using high-level ab initio
calculations.
The reoptimized parameter set has been validated against experimentally observed
conformational distributions for a considerable number of modified residues. We have also
shown that for some of the modifications, like 2-thiolation of uridines and pseudouridylation,
the modifications stabilize the RNA structure. A major role in such stabilization is played by
the enhancement in stacking interaction.
The results obtained suggest that the combination of AMBER force field for RNA with our
reoptimized set of dihedral parameters and the recalculated partial charges provide a better
agreement with experimental data and as such should be more useful in producing a more
accurate model of the structure and function of modified RNA.