A tree-based algorithm for determining the effects of solvation on the structure of salivary gland tripeptide NH3+-D-PHE-D-GLU-GLY-COO-

Abstract

A D-enantiomeric analog of the submandibular gland rat-1 tripeptide FEG (Seq: NH(3)(+)-Phe-Glu-Gly-COO(-)) called feG (Seq: NH(3)(+)-D-Phe-D-Glu-Gly-COO(-)) was examined by molecular dynamics simulations in water. Previous in vacuo simulations suggested a conformation consisting predominantly of interactions between the Phe side chain and glutamyl-carboxyl group and a carboxyl/amino termini interaction. The solvated peptide was simulated using two approaches which were compared-a single 400-ns simulation and a "simulation tree." The "tree" approach utilized 45 10-ns simulations with different conformations used as initial structures for given trajectories. We demonstrate that multiple short duration simulations are able to describe the same conformational space as that described by longer simulations. Furthermore, previously described in vacuo interactions were confirmed with amendments: the previously described head-to-tail arrangement of the amino and carboxyl termini, was not observed; the interaction between the glutamyl carboxyl and Phe side chain describes only one of a continuum of conformations present wherein the aromatic residue remains in close proximity to the glutamyl carbonyl group, and also interacts with either of the two available carboxyl groups. Finally, utilizing only two separate 10-ns trajectories, we were able to better describe the conformational space than a single 60-ns trajectory, realizing a threefold decrease in the computational complexity of the problem.

FIGURE 1

A graphical representation of the simulation tree. The tree as displayed is incomplete. Each bar represents a 10-ns simulation period. The bars to the right of the seed trajectory represent starting structures which exhibited the greatest deviation as compared to the start of the seed run. Bars on the left represent starting structures which were least deviated from the start structure of the seed run. The image does not show an additional set of five simulations originating from the seed run which were started with structures that represented the average of the seed run. Furthermore, each of these branch runs possesses two additional runs denoted by an A or B appended to their source trajectories, representing least deviated and most deviated structures with respect to the original seed structure.

FIGURE 2

The backbone root mean square deviation (RMSD) of feG during the 400-ns, extended duration trajectory as compared to the seed structure. The RMSD is with relation to the backbone position. Although quite compressed, it is clearly evident that feG spends the greatest part of the simulation within a very narrow range, between a backbone RMSD of 0.05 nm and 0.10 nm.

FIGURE 3

Representative backbone root mean square deviation (RMSD) for trajectories S0, S0A, and S0B. Each trajectory is 10 ns in duration. Although some fluctuation occurs, it is evident in these traces that feG spends the majority of the simulation within a narrow range of conformations found between 0.05 nm and 0.10 nm. This range of conformations is what is classified as cluster 1 (Fig. 4).

FIGURE 4

Stereoscopic image of feG in cluster 1. The central structure for each trajectory's cluster 1 has been overlayed. The high degree of similarity over the extent of the backbone is immediately apparent. The only differences between these structures is the rotation of each side chain about χ₁, χ₂, and χ₃. However, even these rotations are minor in comparison to the similarity between the structures. Note the close proximity of the Glu carbonyl to the Phe ring. This is similar to previous observations resulting from in vacuo simulations. However, in solvated simulations the aromatic side chain is able to take a conformation which involves either an interaction with the Glu-carboxyl, as in in vacuo, or with the Gly-carboxyl.

FIGURE 5

Percentage of cluster occurrence for all simulations. In light gray is the occurrence of each cluster over the full 400-ns duration. In dark gray is the aggregated results of the tree simulation, representing a total time of 450 ns. As can be seen, both sets of simulations yielded nearly identical occurrences of any given cluster. Only clusters above number 5 possessed >1% of the cluster distribution. All remaining clusters are observed <1% of the time in any given simulation.

FIGURE 6

A comparison of each trajectory and the number of clusters identified. The length of each bar corresponds to the total number of identified clusters, whereas each of the stacked components represents the percentage of a given trajectory found in a particular cluster. The percentage displayed in each bar is the occurrence of cluster 1 and is given for reference. Clusters increase from left to right, with 1 at left through the maximum for each trajectory.

FIGURE 7

The overlayed backbone of all feG cluster 1 central structures. The measurements of the backbone are summarized in Table 2. Note the very high degree of overlap between all structures. This backbone is likely the necessary scaffold structure required for a biologically active compound.