Adsorption, folding, and packing of an amphiphilic peptide at the air/water interface.

Peptide oligomers play an essential role as model compounds for identifying key motifs in protein structure formation and protein aggregation. Here, we present our results, based on extensive molecular dynamics simulations, on adsorption, folding, and packing within a surface monolayer of an amphiphilic peptide at the air/water interface. Experimental results suggest that these molecules spontaneously form ordered monolayers at the interface, adopting a ß-hairpin-like structure within the surface layer. Our results reveal that the ß-hairpin structure can be observed both in bulk and at the air/water interface. However, the presence of an interface leads to ideal partitioning of the hydrophobic and hydrophilic residues, and therefore reduces the conformational space for the molecule and increases the stability of the hairpin structure. We obtained the adsorption free energy of a single ß-hairpin at the air/water interface, and analyzed the enthalpic and entropic contributions. The adsorption process is favored by two main factors: (1) Free-energy reduction due to desolvation of the hydrophobic side chains of the peptide and release of the water molecules which form a cage around these hydrophobic groups in bulk water. (2) Reduction of the total air/water contact area at the interface upon adsorption of the peptide amphiphile. By performing mutations on the original molecule, we demonstrated the relative role of key design features of the peptide. Finally, by analyzing the potential of mean force among two peptides at the interface, we investigated possible packing mechanisms for these molecules within the surface monolayer.

Driving forces for adsorption of amphiphilic peptides to the air-water interface.

We have studied the partitioning of amphiphilic peptides at the air-water interface. The free energy of adsorption from bulk to interface was calculated by determining the potential of mean force via atomistic molecular dynamics simulations. To this end a method is introduced to restrain or constrain the center of mass of a group of molecules in a periodic system. The model amphiphilic peptides are composed of alternating valine and asparagine residues. The decomposition of the free energy difference between the bulk and interface is studied for different peptide block lengths. Our analysis revealed that for short amphiphilic peptides the surface driving force dominantly stems from the dehydration of hydrophobic side chains. The only opposing force is associated with the loss of orientational freedom of the peptide at the interface. For the peptides studied, the free energy difference scales linearly with the size of the molecule, since the peptides mainly adopt extended conformations both in bulk and at the interface. The free energy difference depends strongly on the water model, which can be rationalized through the hydration thermodynamics of hydrophobic solutes. Finally, we measured the reduction of the surface tension associated with complete coverage of the interface with peptides.

The introduction of hydrogen bond and hydrophobicity effects into the rotational isomeric states model for conformational analysis of unfolded peptides.

Relative contributions of local and non-local interactions to the unfolded conformations of peptides are examined by using the rotational isomeric states model which is a Markov model based on pairwise interactions of torsion angles. The isomeric states of a residue are well described by the Ramachandran map of backbone torsion angles. The statistical weight matrices for the states are determined by molecular dynamics simulations applied to monopeptides and dipeptides. Conformational properties of tripeptides formed from combinations of alanine, valine, tyrosine and tryptophan are investigated based on the Markov model. Comparison with molecular dynamics simulation results on these tripeptides identifies the sequence-distant long-range interactions that are missing in the Markov model. These are essentially the hydrogen bond and hydrophobic interactions that are obtained between the first and the third residue of a tripeptide. A systematic correction is proposed for incorporating these long-range interactions into the rotational isomeric states model. Preliminary results suggest that the Markov assumption can be improved significantly by renormalizing the statistical weight matrices to include the effects of the long-range correlations.