We report here the results on conformational free energies of N-acetyl-N‘-methylprolineamide (Ac-Pro-NHMe) at the HF, MP2, and B3LYP levels of theory with the 6-31+G(d) basis set to see how well the ab initio and density functional computations with the self-consistent reaction field (SCRF) method would reproduce the observed solvent effects on the cis−trans isomerization of imide bond. The augmentation of the prolyl amide cis-isomer population with increased solvent polarity was reproduced by the MP2 or B3LYP free energies using the SCRF solvation free energies. Furthermore, the energy barriers for amide isomerization that were calculated for Ac-Pro-NHMe in different solvents reflected the experimentally observed trend for dimethylacetamide in which the energy for rotation augmented with increased solvent polarity. However, it should be noted that neither the isodensity polarizable continuum model (IPCM) nor the self-consistent isodensity polarizable continuum model (SCI-PCM) predicted properly both solvent effects. Our results indicate the need for improved solvation theory.

A conformational study on the terminally blocked proline oligopeptides, Ac-(Pro)(n)()-NMe(2) (n = 2-5), is carried out using the ab initio Hartree-Fock level of theory with the self-consistent reaction field method in the gas phase and in solutions (chloroform, 1-propanol, and water) to explore the preference and transition between polyproline II (PPII) and polyproline I (PPI) conformations depending on the chain length, the puckering, and the solvent. The mean differences in the free energy per proline of the up-puckered conformations relative to the down-puckered conformations for both diproline and triproline increases for the PPII-like conformations and decreases for the PPI-like conformations as the solvent polarity increases. These calculated results indicate that the PPII-like structures have preferentially all-down puckerings in solutions, whereas the PPI-like structures have partially mixed puckerings. The free energy difference per proline residue between the PPII- and PPI-like structures decreases as the proline chain becomes longer in the gas phase but increases as the proline chain becomes longer in solutions and the solvent polarity increases. In particular, our calculated results indicate that each of the proline oligopeptides can exist as an ensemble of conformations with the trans and cis peptide bonds in solutions, although the PPII-like structure with all-trans peptide bonds is dominantly preferred, which is reasonably consistent with the previously observed results. In diproline Ac-(Pro)(2)-NMe(2), the rotational barrier to the cis-to-trans isomerization for the first prolyl peptide bond increases as the solvent polarity increases, whereas the rotational barrier for the second prolyl peptide bond does not show the monotonic increase as the solvent polarity increases. When the rotational barriers for these two prolyl peptide bonds were compared, it could be deduced that the conformational transition from PPI with the cis peptide bond to PPII with the trans peptide bond is initiated at the C-terminus and proceeds to the N-terminus in water. This is consistent with the results from NMR experiments on polyproline in D(2)O but opposite to the results from enzymatic hydrolysis kinetics experiments on polyproline.

The conformational study on L-azetidine-2-carboxylic acid (Ac-Aze-NHMe, the Aze dipeptide) and (S)-piperidine-2-carboxylic acid (Ac-Pip-NHMe, the Pip dipeptide) is carried out using ab initio HF and density functional methods with the self-consistent reaction field method to explore the differences in conformational preferences and cis-trans isomerization for proline residue and its analogues with different ring size in the gas phase and in solution (chloroform and water). The change of ring size by deleting a CH2 group from or adding a CH2 group to the prolyl ring results the remarkable changes in backbone and ring structures compared with those of the Pro dipeptide, especially in the C'-N imide bond length and the bond angles around the N-C(alpha) bond. The four-membered azetidine ring can have either puckered structure depending on the backbone structure because of the less puckered structure. The six-membered piperidine ring can adopt chair and boat conformations, but the chair conformation is more preferred than the boat conformation. These calculated preferences for puckering are consistent with experimental results from analysis of X-ray structures of Aze- and Pip-containing peptides. On going from Pro to Aze to Pip, the axiality (i.e., a tendency to adopt the axial orientation) of the NHMe group becomes stronger, which can be ascribed to reduce the steric hindrances between 1,2-substituted Ac and NHMe groups. As the solvent polarity increases, the polyproline II-like conformation becomes more populated and the relative stability of conformation tC with a C7 hydrogen bond between C'=O of the amino group and N-H of the carboxyl group decreases for both the Aze and Pip dipeptides, as seen for the Pro dipeptide. The cis population and rotational barriers for the imide bond increase with the increase of solvent polarity for both the Aze and Pip dipeptides, as seen for the Pro dipeptide. In particular, the cis-trans isomerization proceeds in common through only the clockwise rotation with omega' approximately +120 degrees about azetyl and piperidyl peptide bonds in the gas phase and in solution, as seen for alanyl and prolyl peptide bonds. The pertinent distance d(N...H-N(NHMe)) and the pyramidality of imide nitrogen can describe the role of this hydrogen bond in stabilizing the transition state structure, but the lower rotational barriers for the Aze and Pip dipeptides than those for the Pro dipeptide, which is observed from experiments, cannot be rationalized.

The conformational study on 20 Ac-Xaa-Pro-NHMe dipeptides has been carried out using an empirical potential function ECEPP/3 in order to investigate the factors responsible for the preference of proline puckering of the peptides with the trans or cis imide bond preceding the proline. The general conformational preference for down- and up-puckered dipeptides is calculated as trans-down > trans-up > cis-down > cis-up, which is reasonably in accord with that estimated by analyzing X-ray structures of proteins and the result for the single proline residue. The overestimated occurrence of trans-down conformations of proline seems to be caused by excluding long-range interactions that short dipeptides cannot have. The average computed occurrence of dipeptides with cis imide bonds is about 3%, somewhat lower than the value calculated for Ac-Pro-NHMe, which is close to experimental estimates obtained from X-ray structures of proteins. In particular, the interaction of the aromatic side chain of Xaa residue with the proline ring appears not to be strong enough to stabilize the stacked conformations of small dipeptides with cis imide bonds. The propensity to adopt trans or cis imide bond and to form secondary structures of Xaa-Pro sequences is discussed and compared with results obtained from X-ray structures of proteins.

Conformational study of RGD tripeptides in the nonhydrated and hydrated states was carried out using an empirical potential function ECEPP/3 and the hydration shell model in order to investigate preferred conformations and factors responsible for their stability. RGD tripeptides in the nonhydrated and hydrated states can be interpreted as existing as an ensemble of feasible conformations rather than as a single dominant conformation from the analysis of distributions of backbone conformations, hydrogen bonds and beta-turns. The different distributions of conformations for the neutral and zwitterionic RGD tripeptides in both states may indicate that the conformation of the RGD tripeptide is liable to depend on solvent polarity and pH values. beta-Turn populations for the neutral tripeptide in both states are reasonably consistent with NMR measurements on linear RGD-containing peptides. The degradation of RGD tripeptide seems to be attributed mainly to the hydrogen bonds between the Asp side-chain and the backbone of Asp residue or C-terminal NHMe group, rather than to the flexible backbones of Gly and Asp residues.