The (3)J(C8-H1'), (3)J(C4-H1'), (1)J(C8-H8), (1)J(C1'-H1'), (1)J(C2'-H2'), and (1)J(C2'-H2'2) indirect scalar coupling constants were calculated with the density functional theory in the deoxyguanosine and riboguanosine molecules. The following geometry descriptors were considered in analysis of the structural dependence of the six J couplings: the glycosidic torsion angle chi and conformation of the hydroxymethyl group at the C4' carbon of sugar mimicking the backbone residue and the sugar pucker (C2'-, C3'-endo). The (3)J(C8-H1') and (3)J(C4-H1') couplings, which are typically assigned to the chi torsion, also depended on the sugar pucker, although the calculated dependence of the latter coupling on sugar pucker was nearly negligible. New parametrization of the Karplus equations, taking into account the stereoinversion effect at the glycosidic nitrogen atom and solvent effects, was calculated for the (3)J(C8-H1') and (3)J(C4-H1') coupling assigned to the chi torsion. The calculated phase shift of chi torsion angle in these new Karplus equations was larger by approximately 10 degrees compared to its commonly accepted value of 60 degrees (Wijmenga, S. S.; van Buuren, B. N. M. Prog. NMR Spectrosc. 1998, 32, 287.). The calculated (1)J(C2'-H2') and (1)J(C2'-H2'2) coupling dominantly depended on the sugar type (deoxyribose or ribose) and its pucker, while the (1)J(C1'-H1') and (1)J(C8-H8) coupling dominantly depended on the glycosidic torsion angle, although quantitatively, all four (1)J couplings depended on both geometry parameters. The dependences of j-couplings on the torsion angle chi calculated in isolated nucleosides were compared with those taking into account the effect of base pairing occurring in the WC/SE RNA base pair family, which appeared to be minor. The calculated (3)J couplings agreed well with available experimental data similarly as the (1)J couplings, although lack of experimental data diminished more reliable validation of the later couplings.
Read full abstract