Toward a unified picture of nucleic acid dynamics: Modeling results from both solution and solid state NMR using realistic potentials

Nucleic acid dynamics compel interest, both for their biological relevance and for the many physical unknowns about the details of their motions. Despite a growing body of experimental work investigating these dynamics by means of solution and solid state NMR spectroscopies, until now there have been few efforts to a develop a unified theoretical framework for the quantitative comparison of results from the two techniques. This thesis outlines such an approach, and applies it to two nucleic acid systems of interest. Binding sites in the Hha1 methyltransferase target DNA dodecamer and the HIV-1 transactivation response (TAR) RNA recognition site element both exhibit significant rearrangements critical to function. Because solid state lineshapes strongly constrain the choice of model parameters, for both biological systems the solid state NMR parameters are fit first. The attempt is made to predict solution parameters assuming the same internal motions as in solids. In the case of the fairly rigid DNA fragment this procedure is successful. In the case of the RNA 29-mer, fitting the solid state parameters holds intrinsic interest. However it does not prove accurate to assume that rigid tumbling of these motions can reproduce solution motions.