2-Deoxy-beta-D-erythro-pentofuranose: hydroxymethyl group conformation and substituent effects on molecular...

The effect of hydroxymethyl conformation (gg, gt, and tg rotamers about the C4-C5 bond) on the conformational energies and structural parameters (bond lengths, bond angles, bond torsions) of the 10 envelope forms of the biologically relevant aldopentofuranose, 2-deoxy-beta-D-erythro-pentofuranose (2-deoxy-D-ribofuranose) 2, has been investigated by ab initio molecular orbital calculations at the HF/6-31G level of theory. C4-C5 bond rotation induces significant changes in the conformational energy profile of 2 (2gt and 2tg exhibit one global energy minimum, whereas 2gg exhibits two nearly equivalent energy minima), and structural changes, especially those in bond lengths, are consistent with predictions based on previously reported vicinal, 1,3- and 1,4-oxygen lone pair effects. HF/6-31G-optimized envelope geometries of 2gg were re-optimized using density functional theory (DFT, B3LYP/6-31G), and the resulting structures were used in DFT calculations of NMR spin-spin coupling constants involving 13C (i.e., J(CH) and J(CC) over one, two, and three bonds) in 2gg according to methods described previously. The computed J-couplings were compared to those reported previously in 2gt to assess the effect of C4-C5 bond rotation on scalar couplings within the furanose ring and hydroxymethyl side chain. The results confirm prior predictions of correlations between 2J(CH), 3J(CH), 2J(CC) and 3J(CC), and ring conformation, and verify the usefulness of a concerted application of these couplings (both their magnitudes and signs) in assigning preferred ring and C4-C5 bond conformations in aldopentofuranosyl rings. The new calculated J-couplings in 2gg have particular relevance to related J-couplings in DNA (and RNA indirectly), where the gg rotamer, rather than the gt rotamer, is observed in most native structures. The effects of two additional structural perturbations on 2 were also studied, namely, deoxygenation at C5 (yielding 2,5-dideoxy-beta-D-erythro-pentofuranose 4) and methyl glycosidation at O1 (yielding methyl 2-deoxy-beta-D-erythro-pentofuranoside 5) at the HF/6-31G level. The conformational energy profile of 4 resembles that found for 2gt, not 2gg, indicating that 4 is an inappropriate structural mimic of the furanose ring in DNA. Glycosidation failed to induce differential stabilization of ring conformations containing an axial C1-O1 bond (anomeric effect), contrary to experimental data. The latter discrepancy indicates that either the magnitude of this differential stabilization depends on ring configuration or that solvent effects, which are neglected in these calculations, play a role in promoting this stabilization.