Conformational Variability of RNA Backbone

Bohdan Schneider1, Zdenek Moravek2, and Helen M. Berman3

1Center for Complex Molecular Systems and Biomolecules, Dolejskova 3, CZ-18223 Prague, Czech Republic, bohdan.schneider@jh-inst.cas.cz

2Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic

3Rutgers University, Piscataway, NJ-08854, USA.

 

As shown by ribozyme and especially ribosome structures solved in last few years, molecules of RNA form complicated 3D folds which have no match among known DNA structures but their complexity is quite comparable to that of protein folds. Complicated RNA folds are enabled by a high flexibility of the nucleotide backbone but little is known about its conformational behavior. A well refined structure of the large ribosome subunit 50S at 2.4A, NDB structure RR0033 (PDB ID 1JJ2), Ban et al. Science 289, 905 (2000), provides a database of over 2700 nucleotides. This work analyzes conformations of these nucleotides by a combination of Fourier averaging and clustering techniques.

Majority of all nucleotides of RR0033, about 70%, are in the A-type conformation, this main conformational type can be further classified into three subclasses. The remaining 30% of nucleotides with other than A-type conformations were analyzed in a greater detail. The backbone torsion angles for each nucleotide were grouped into eight sets of three angles with the main emphasis on the torsions around the two phosphodiester bonds, O3*-P (torsion zeta) and P-O5* (alpha). Each set of three torsions results in a 3D distribution of points in a parametric torsional space and this distributions was Fourier transformed into densities of nucleotide conformations. Peak positions (maxima) of these maps confine the most probable (di)nucleotide conformations.

Nucleotides belonging to the same peaks in several torsional 3D maps have similar geometry. Such nucleotides were grouped and compared in Cartesian (real) 3D space. In such a way, twelwe types of  highly untypical (non-A) nucleotide conformations were identified and their Cartesian coordinates determined. These untypical nucleotide conformations can be useful in e.g.refinement process and are available upon request.

 

BS is grateful to support by grant LN00A032 from the Ministry of Education of the Czech Republic.