Computational study into the stabilizing interactions in internal loops of oligonucleotides
Our focus in this computational proposal is on weak bonding interactions (WBIs) in the organization of molecular assemblies, here in particular oligonucleotides. With regards to WBIs, the hydrogen bond is certainly the most recognized, but C-H…O, C-H…pi, pi…pi and others must not be ignored if organization and stability are to be understood. WBIs are thus of fundamental importance in diverse phenomena, such as the effect of ligand-binding in ribonucleic acids on gene expression (riboswitches, mostly through pi…pi interactions). Our approach is to compute -mostly using density-functional theory methods- and analyze electron densities to identify and predict WBIs in oligonucleotides. RNA structure prediction from scratch is still an elusive goal, largely due to ill-defined loop regions. From available structures (crystal and NMR), we will rigorously characterize the network of WBIs in selected ill-defined RNA secondary and tertiary structures to develop a database of recurring WBIs for later use in RNA structure prediction.
In earlier studies on nucleic acids we have shown (Journal of Physical Chemistry A 115, 12800-12808 (2011)) that a fully optimized, planar base pair is a poor model of the base-pairing situation as found in X-ray and NMR structures of oligonucleotides, where the relative orientation of a base to its surrounding bases is dictated by the stability of the whole structure. Base pair geometries are distorted from their optimum planarity, and this leads to changes in the hydrogen bonding network. Changes in the relative orientation of a base pair with respect to another leads to changes in pi-stacking.
The kind of detailed insight that can be gained from studies on WBIs in oligonucleotides is illustrated in the fact that we were able to solve a long-standing controversy, namely whether AT or AU base pairs are more stable. Experimental evidence supports both views, and our work suggests that this seeming discrepancy is grounded in the particular choice of nucleic acids in the different studies. We found the hydrogen-bond densities, and therefore strengths, to vary widely, and any experimental conclusion therefore must strongly depend on the choice of system. Only a study on a large number of systems as ours reveals that, on average, there is no difference.
We now wish to use our proven methodology to determine WBIs in structurally ill-defined internal loops, with the ultimate goal to aid in the development of improved structure-determination algorithms. Candidates for our study will be selected oligonucleotides that form a hairpin loop or a duplex with internal bulges.
