Secondary structure of the hybrid poly(rA) · poly(dT) in solution: Studies involving NOE at 500 MHz and stereochemical modelling within the constraints of NOE data

Abstract

One-dimensional nuclear Overhauser effect (NOE) in nuclear magnetic resonance spectroscopy along with stereochemically sound model building was employed to derive the structure of the hybrid poly(rA) · poly(dT) in solution. Extremely strong NOE was observed at AH2' when AH8 was presaturated; strong NOEs were observed at TH2′TH2′′ when TH6 was presaturated; in addition the observed NOEs at TH2′ and TH2′′ were nearly equal when TH6 was presaturated. There was no NOE transfer to AH3′ from AH8 ruling out the possibility of (C-3′- endo, low anti χ ∼- 200 ° to 220 °) conformation for the A residues. The observed NOE data suggest that the nucleotidyl units in both rA and dT strands have equivalent conformations: C-2′- endo/C-1′- exo, anti χ ∼- 240 ° to 260 °. Such a nucleotide geometry for rA/dT is consistent with a right-handed B-DNA model for poly(rA) · poly(dT) in solution in which the rA and dT strands are conformationally equivalent. Molecular models were generated for poly(rA) · poly(dT) in the B-form based upon the geometrical constraints as obtained from the NOE data. Incorporation of (C-2′- endo pucker, χ ∼- 240 ° to 260 °) into the classical β-form resulted in severe close contacts in the rA chain. By introducing base-displacement, tilt and twist along with concomitant changes in the backbone torsion angles, we were able to generate a β-form for the hybrid poly(rA) · poly(dT) fully consistent with the observed NOE data. In the derived model the sugar pucker is C-1′- exo, a minor variant of C-2′- endo and the sugar base torsion is 243 °, the remaining torsion angles being: ϵ = 198 °, ξ = 260 °, α = 286 °, β = 161 ° and γ = 72 °; this structure is free of any steric compression and indicates that it is not necessary to switch to C-3′- endo pucker for rA residues in order to accommodate the 2′-OH group. The structure that we have proposed for the polynucleotide RNA — DNA hybrid in solution is in complete agreement with that proposed for a hexamer hybrid in solution from NOE data and is inconsistent with the heteronomous model proposed for the fibrous state.