Dynamical studies of peptide motifs in the Plasmodium falciparum circumsporozoite surface protein by restrained and unrestrained MD simulations.

The immunodominant region on the circumsporozoite surface (CS) protein of the malaria parasite Plasmodium falciparum contains 37 repeated copies of a asparagine-alanine-asparagine-proline (NANP) motif NMR studies of linear synthetic peptides containing one, two or three repeat units provided evidence for nascent type I beta-turns within the NPNA cadence in aqueous solution. The beta-turns could be stabilised upon substituting proline for alpha-methylproline (p(Me)) in the dodecamer (NP(Me)NA)3, without loss of the ability to elicit antibodies cross-reactive with P. falciparum sporozoites. In this work, four 4 ns MD simulations of the dodecapeptide Acetyl-(NP(Me)NA)3, in water, using NOE distance restraints, using 3J-coupling constant restraints, using both these restraints and without restraints, were carried out to determine the conformations of this peptide in aqueous solution. An unrestrained MD simulation of the unmethylated Ac-(NPNA)3 peptide in water was also carried out to investigate the effect of the additional methyl groups on the structure and dynamics of the peptide. The application of NOE distance restraints and 3J-coupling constant restraints leads to contradictory results, probably due to different averaging time scales inherent to the measurement of these data, which exceed the 100 ps averaging applied in the simulations. The additional methyl groups lead to more compact structures, which display enhanced local fluctuations. The central tetrapeptide adopts a type I beta-turn, while the outer motifs display more conformational variability. The three motifs in the methylated dodecamer peptide, however, adopt frequently in the distance restrained MD simulation a compact structure such that the outer motifs appear to form a hydrophobic core by stacking of their two proline rings. This arrangement also suggests how a peptide containing multiple tandemly linked copies of a stable beta-turn NPNA motif might adopt a folded stem-like structure, which conceivably may be of biological relevance in the native CS protein.

Molecular dynamics simulation using weak-coupling NOE distance restraining.

Application of the weak-coupling scheme to restrain the configurations of a molecular system to a set of NOE distance restraints is investigated using two test systems: (i) a 15-atom chain molecule with one distance restraint; and (ii) a protein molecule with hundreds of NOE distance restraints. Atom-atom distance restraining by the weak-coupling technique is possible, but this method does not produce as good results as the penalty function method normally used to maintain NOE distance restraints.

Parametrisation of time-averaged distance restraints in MD simulations.

Time-averaged restraints in molecular dynamics simulations offer a means to account for the averaging that is implicit in NMR spectroscopic data. We present a systematic investigation of the parameters which characterise time-averaged distance restraints. Using previously published data for a small protein, chymotrypsin inhibitor 2, we identify conditions which can lead to undesirable heating or which grossly distort the dynamics of the system.

A reassessment of the structure of chymotrypsin inhibitor 2 (CI-2) using time-averaged NMR restraints.

Chymotrypsin inhibitor 2 (CI-2) is one of the growing family of proteins for which well-defined solution and crystal structures have been published and for which small, but distinct differences between these were found. It presents an ideal case to address the question of whether a structural difference is physically real or due to the simplifying approximations with respect to averaging that are used in the conventional methods for structure refinement. NOE distance and 3J coupling constant restrained molecular dynamics simulations were performed using conventional and time-averaged restraints, both in vacuo and in aqueous solution, and the trajectories were compared with structural properties of published structures. The time-averaged restrained molecular dynamics simulations sampled more conformations at various times and visited states consistent with both previously published solution and crystal structures. It was found that the difference between these structures is due to the refinement methodology used. Application of time-averaged restraints in structure refinement yields a physically different picture of the molecular mobility.