Journal of Physical Chemistry A: Molecules, Spectroscopy, Kinetics, Environment and General Theory 2009-10-08

Quantum mechanical and NMR studies of ring puckering and cis/trans-rotameric interconversion in prolines and hydroxyprolines.

Abil E Aliev, Simrath Bhandal, Denis Courtier-Murias

文献索引：J. Phys. Chem. A 113(40) , 10858-65, (2009)

全文：HTML全文

摘要

Nuclear magnetic resonance (NMR) and quantum mechanical (QM) studies have been carried out for proline (Pro) containing peptides: N-acetyl-l-proline (AcProOH) and N-acetyl-4-hydroxy-l-proline (AcHypOH). Preliminary results of variable temperature NMR measurements for Gly-Pro-Gly-Gly (GPGG), Val-Ala-Pro-Gly (VAPG), and Ala-Pro-Gly-Trp amide acetate salt (APGW) are also reported. The effect of solvent (D(2)O, DMSO-d(6) and CD(3)CN) on the pyrrolidine ring conformation and cis/trans-rotamerisation along the amide bond preceding Pro was investigated by temperature dependent NMR followed by detailed transition state (TS) searches for both conformational equilibria using QM methods. The results revealed the energetic characteristics of the TS, which were in satisfactory agreement with NMR, and the corresponding TS geometries, which are not available from experiment. The most remarkable feature of the cis/trans-rotamerisation is that the amide nitrogen in AcProOH and AcHypOH adopts a tetrahedral geometry in the TS. Various HF, DFT, and MP2 calculations together with implicit solvation modeling were employed in order to identify the most suitable QM protocols for reliable predictions of the geometry and the relative energies of the conformations of Pro and Hyp containing peptides in aqueous solution. Solution NMR results were used for the verification of the reliability of the QM predictions. The results indicate that the MP2 calculations combined with implicit solvation models are reasonably accurate in reproducing NMR measured populations of four different conformations of either AcProOH or AcHypOH in different solvents, whereas HF and DFT B3LYP calculations were significantly less accurate.