Info

Sensitivity of the Folding Equilibrium to the Force Field Used

Most of the simulation studies of the folding equilibria of peptoids under various thermodynamic and environmental conditions made use of the GROMOS force field [47, 48], parameter sets 43A1 [49] or 45A3 [50]. These parameter sets contain nonbonded interaction parameters for nonpolar atoms which were optimized to reproduce thermodynamic properties (heat of vaporization, density, free energy of solvation) for liquid hydrocarbons and their aqueous solutions [49, 50]. Because of this thermodynamic basis, the various predominant folds of the different b-peptides with predominantly nonpolar side chains could be found in the MD trajectories in agreement with NMR experimental data [11-13, 16-19]. There seem to be no comparable studies based on other biomolecular force fields that show nearly as good agreement with the experimental data as is obtained using the GROMOS force field [51]. However, the nonbonded interaction parameters for polar atoms in the 43A1 and 45A3 GROMOS force fields had not yet been optimized to reproduce the above-mentioned thermodynamic properties for liquids of polar molecules and their aqueous solutions. Such an optimization led to the GROMOS 53A6 force-field parameter set [48]. It came as no surprise that a simulation using the 45A3 GROMOS force field for a 12-b-peptide with predominantly polar side chains could not reproduce the 314-helix experimentally observed to be stable in methanol. Only with the 53A6 force-field parameters this helical fold became stable [39], as is illustrated in Fig. 6.11. For the 20-b-peptide

Fig. 6.11 Atom-positional root-mean-square deviation of the backbone atoms of residues 2-11 (the structure of the peptide is given in the figure) with respect to the experimental NMR model structure derived for the peptide in methanol. Parameter sets 45A3 (black) and 53A6 (red) in methanol [39].

Fig. 6.11 Atom-positional root-mean-square deviation of the backbone atoms of residues 2-11 (the structure of the peptide is given in the figure) with respect to the experimental NMR model structure derived for the peptide in methanol. Parameter sets 45A3 (black) and 53A6 (red) in methanol [39].

6.6 Comparison of Simulated with Experimentally Measured Observables | 185

in solution (Fig. 6.4), the 53A6 force field seems to preserve the 314-helix slightly better than the 45A3 force field, but overall the picture is essentially the same for this molecule.

Not surprisingly, a calibration of force-field parameters for small molecules that represent the various moieties present in peptide analogs against thermodynamic data in the condensed phase seems a necessary condition to adequately simulate the folding equilibria of polypeptide analogs in various solvents. Not only the sol-vation properties in aqueous solution should correspond to experimental data, but also those for other solvents such as chloroform, cyclohexane, methanol, DMSO, acetronitrile and acetone [52], which may be used as solvents for bio-molecular studies.

0 0

Post a comment