This overall project focuses on the study of membranes, proteins and carbohydrates by molecular dynamics computer simulation. Progress is reported under each Aim listed above Aim 1. Understand Model Membranes. Paper 1 did not involve membranes, but did involve a model polymer, polyethylenimine (PEI). PEI has the chemical structure CH3-(CH2-NH2+-CH2)n -CH3. It is a logical extension of linear polymers studied previously in this project, alkanes, CH3-(CH2-CH2-CH2)n -CH3, and polyethenlene glycol, CH3-(CH2-O-CH2)n -CH3. In this case, complexity is increased because the amine group is protonable. PEI is used extensively in applications which take advantage of the polymers high charge density and the pH-dependence of its charge, including metal chelation, drug delivery, enzyme-trapping, and antimicrobial surfaces. The present project examined radius of gyration of PEI at infinite dilution as a function of charge using simulation and experiment, and showed that protonation dependent persistence length is required to obtained agreement with analytical theory. The role of mannobiosylation was also considered, and hydrodynamic estimates were used to infer the presence of aggregation. Constant pH simulations are of interest for the field, and PEI might provide a model system. Paper 2 involved revisiting techniques for extracting volumes for the component groups in lipid bilayers (e.g., methylene, methyl and head groups). The traditional approach using only experimental data had led to the counterintuitive results that head group volumes in bilayers were deceasing with increasing temperature. A new analysis utilizing simulations and more rigorous connection to partial molar volumes resolved the controversy, and allowed the definition of a new quantity: component free volumes. Component free volumes are calculated by subtracting the hard core excluded volume obtained from the Lennard-Jones radii for a particular group from its component volume. Hence, the free volume is obtained for different regions of a bilayer, which lends insight to permeabilities in the different phases. In the coming year the approach will be extend to liquid ordered phases.
Aim 2. Develop Simulation Methodology. Paper 3 considered a number of topics related to the evaluation of the membrane permeability of different solutes using computer simulation. While some is review, an original method for evaluating permeability by counting transition was developed and applied to water and oxygen. The results were compared with those from Bayesian analysis methods reported last year. The counting method was used compare the permeabilities of water and oxygen in liquid ordered and liquid disordered membranes in a paper that was submitted this year, but has not yet been accepted. Another finding reported in Paper 3, that polarizability in a critical factor in the permeability of polar molecular such as water, was further developed in Paper 4. This paper shows that effects of polarizability in membrane permeation can be modelled in an inexpensive manner within an additive force field (FF): modification of the Lennard-Jones interaction terms between the permeant and the hydrocarbons. It was demonstrated that this modification dramatically improves that agreement of water permeation for a set of bilayers. While using an explicitly polarizability FF such as the Drude is arguably a better solution in this case, the method could be useful for cases where polarizability in critical for only a small part of the system (e.g., a probe in a bilayer).
Aim 3. Simulate Complex Membranes Paper V returned to antimicrobial peptides (AMP) piscidins 1 and 3. The major contribution of simulations to this multi-method study was to reconcile the results at high hydration, where the peptides are surface bound on average, and low hydration, where they are deeply inserted. The simulations showed that the fluctuations at high hydration lead to deeply inserted conformations that are, in fact, consistent with the insertion angles obtained at low hydration. These results support a defect model of AMP action and further disputes the pore hypothesis. Work is underway in the to develop more efficient sampling methods for peptide conformation in membranes and to simulate pH dependence. The experimental results present in Paper V will be invaluable benchmarks for this work.
Showing the most recent 10 out of 52 publications
