Prediction of molecular alignment of nucleic acids in aligned media

Abstract

We demonstrate – using the data base of all deposited DNA and RNA structures aligned in Pf1-medium and RDC refined – that for nucleic acids in a Pf1-medium the electrostatic alignment tensor can be predicted reliably and accurately via a simple and fast calculation based on the gyration tensor spanned out by the phosphodiester atoms. The rhombicity is well predicted over its full range from 0 to 0.66, while the alignment tensor orientation is predicted correctly for rhombicities up to ca. 0.4, for larger rhombicities it appears to deviate somewhat more than expected based on structural noise and measurement error. This simple analytical approach is based on the Debye–Huckel approximation for the electrostatic interaction potential, valid at distances sufficiently far away from a poly-ionic charged surface, a condition naturally enforced when the charge of alignment medium and solute are of equal sign, as for nucleic acids in a Pf1-phage medium. For the usual salt strengths and nucleic acid sizes, the Debye–Huckel screening length is smaller than the nucleic acid size, but large enough for the collective of Debye–Huckel spheres to encompass the whole molecule. The molecular alignment is then purely electrostatic, but it’s functional form is under these conditions similar to that for steric alignment. The proposed analytical expression allows for very fast calculation of the alignment tensor and hence RDCs from the conformation of the nucleic acid molecule. This information provides opportunities for improved structure determination of nucleic acids, including better assessment of dynamics in (multi-domain) nucleic acids and the possibility to incorporate alignment tensor prediction from shape directly into the structure calculation process. The procedures are incorporated into MATLAB scripts, which are available on request.

Supporting information available:

Table showing comparison of predicted and experimental magnitude of alignment tensor and two Tables and Figures showing comparison of alignment tensor prediction via method presented here with prediction via ‘steric’ Pales and ‘electrostatic’ Pales. Complete set of correlation plots of displaying predicted vs. experimental RDCs of RNAs/DNAs in Table 1. Derivation of analytical expressions for the electrostatic alignment presented here based on the Debye–Huckel approximation of the electrostatic potential.

The supporting information is available in electronic format at http://dx.doi.org/10.1007/s10858-006-9008-y.