Peptide aggregation and solvent electrostriction in a simple zwitterionic dipeptide via molecular dynamics simulations

Abstract

We investigate the structure of the glycyl-l-alanine dipeptide in aqueous solution at a 1:20 peptide:water concentration via classical, atomistic molecular dynamics simulations using the CHARMM22 force field, and compare to recent neutron diffraction data [S. E. McLain, A. K. Soper, and A. Watts, Eur. Biophys. J. 37, 647 (2008); S. E. McLain, A. K. Soper, I. Diadone, J. C. Smith, and A. Watts, Angew. Chem. Int. Ed. 47, 9059 (2008)]. Comparison between simulations and experiments is made using the static structure factor S(Q). The effect of water model (TIP3P, TIP4P, and SPC/E) upon the solution structure is investigated. Agreement between experiment and simulation is generally good across the entire Q range, although some model-dependent variation is observed, particularly in the predicted intensities of features in S(Q). Peptide aggregation is found to be driven by "hydrophilic" (often bifurcated) hydrogen bonds formed between carboxy and amine functional groups, although simulations suggest that the degree of aggregation is less than that observed experimentally. It is found that hydrophobic association is not significant, with hydrophobic hydration being preferred to association. Detailed examination of the solute structural motifs reveals the existence of bifurcated motifs that are suggested to be an artifact of the CHARMM force field, and may imply that classical force fields provide a flawed structural and dynamical description of such molecular fluids. Investigation of the water structure reveals the presence of an electrostrictive effect which manifests itself as an increase in the number of interstitial molecules in the water second coordination shell, in contradiction to suggestions that this phenomenon arises owing to hydrogen bond bending. Detailed analysis based upon two-dimensional distribution functions suggests an intimate link between the phenomenon of electrostriction and the behavior of water under high-pressure compression. We find the magnitude of the electrostrictive effect inferred from the neutron diffraction data to be greater than that found in the simulations. Investigation of the solvation structure suggests that the CHARMM force field overhydrates the terminal carboxy group, and that this overhydration is accompanied by the presence of bifurcated hydrogen bonds.

Dipeptide Aggregation in Aqueous Solution from Fixed Point-Charge Force Fields

“…In comparison to the EPSR model, the peaks in the radial distribution function are also shifted to larger values. Similar observations were made by others for the nonpolarizable AA-OPLS and CHARMM22 force fields 19,23,24 with different water models, indicating that commonly employed procedures to derive parameters for fixed point-charge force fields are in general not appropriate to describe aggregation of peptides in water. Figure 3 shows that the CHARMM36 force field (c36) in combination with the TIP3P water model leads to results that are in good agreement with the EPSR reference, predicting a somewhat understructured radial distribution function.…”

“…For instance, Tulip and Bates analyzed results from MD simulations with the CHARMM22 22 force field and three different water models and observed less aggregation than derived from the experiment. 23 Similarly, Kucukkal and Stuart 24 found a disappointing correlation with experiments when conducting MD simulations of the Gly-Ala, Gly-Pro, and Ala-Pro dipeptides, with either the polarizable CHARMM30 25,26 or the fixed point-charge CHARMM22 force fields for the dipeptides, in combination with the polarizable TIP4P-FQ 27 and fixed point-charge TIP3P 28 water models, respectively. In the present contribution, we show that a similar problem exists for the AMBER ff12SB force field, which differs from the AMBER ff99SB 18,29 force field only in backbone and side chain torsion parameters but uses the same RESP derived charges.…”

First principles determination of structural, electronic and lattice dynamical properties of a model dipeptide molecular crystal

“…it is interactions between these groups that drive the formation of the crystal structure. It is of interest to note that a recent examination of the structure of the aqueous solution using classical molecular dynamics [47] obtained a similar result, in which molecular aggregation was dominated by charge-charge interactions. Note that in some cases, disagreement between hydrogen bond lengths is marked.…”

“…The hydration structure of the peptide in water has been studied using both classical and ab initio molecular dynamics methods [44,45]. More recently, the solution structure of the peptide in water at a 1:20 solute-solvent concentration has been investigated using neutron diffraction methods [46], whilst a recent study by the current authors [47] has addressed the nature of peptide aggregation in this system and elucidated the solvent structure using classical molecular dynamics simulations. The latter work has provided the stimulus for the current study: molecular crystals provide a well-defined system that may be analysed at a high level of theoretical treatment; in particular, the electronic structure itself can be investigated in some depth, allowing the nature of the intermolecular interactions to be elucidated.…”

Chiral structure of F-actin bundle formed by multivalent counterions

“…We use κ −1 = 10σ. One should note that effects such as electrostriction which may have some contribution [40][41][42] are not considered here. We denote the numbers of F-actins, 3-valent counterions and monomers of each F-actin by N p , N c , and N m , respectively.…”