ANALYSIS OF CONSERVED 3D RNA MOTIFS

Bowling Green State Univ Bowling Green, Bowling Green OH

Investigators

Linked publications & trials

Abstract

Biologically active RNA molecules fold hierarchically to form complex 3- dimensional structures. Folded states of RNA molecules are stabilized by base-upon-base stacking interactions and edge-to-edge, base-pairing, mediated by electrostatically driven hydrogen bonding interactions. In the previous grant period an exhaustive analysis of base-pairing in RNA was carried out. A systematic nomenclature based on geometric considerations was proposed which serves to unambiguously define each base-pairing geometry and to readily identify isosterically related pairs. The value of this approach for identifying recurrent RNA motifs on the basis of sequence comparisons in homologous RNA molecules was demonstrated. The isostericity of the canonical RNA basepairs is the basis of the comparative approach to secondary structure determination. The goal of this proposal is to extend this approach to non-canonical base pairs by determining Isostericity Matricies for each isosteric pairing geometry identified in high- resolution structures, building on the results of the previous grant period. Experimentally determined three-dimensional RNA structures available in the structure databases (NDB and PDB) will be analyzed in light of phylogenetic sequence variations. Potentially isosteric base substitutions will be evaluated by hand-modeling in the experimental structures and by carrying out nano-second molecular dynamics simulations using the AMBER force field with explicit solvent molecules and counterions, and the particle mesh Ewald (PME) method to accurately calculate electrostatic interactions. Selected base substitutions will be incorporated into synthetic oligonucleotides, and their effect on the thermal melting behavior will be measured to provide independent experimental data to compare with computer modeling and simulation. The systematic analysis of base-base stacking patterns in RNA will also be undertaken. Preliminary data indicate that stacking patterns are sequence-dependent and recurrent. Superimposable stacking patterns are seen for particular sequences (for example 5'-AC-3' or 5-UA-3') in different contexts. Moreover, certain sequences exhibit a greater propensity for intra- or cross-strand stacking than others. Intra-strand stacking will first be systematically analyzed by examining all occurrences of each dinucleotide in all available structures. Examples of results this approach has already yielded are presented in the proposal. Inter-strand stacking will then be analyzed. Intra-strand stacking propensities are of particular interest as transient formation of locally stacked bases appears to play an important role in nucleating the early stages of RNA folding, for example in the formation of stable hairpin loops.