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. This first project in this Aim capped a series of papers on gramicidin (gA). This small membrane peptide provides an excellent model of lipid/protein interactions because changes in lipid composition shift its equilibrium from the ion-conducting dimer state to the non-conducting monomer state. Continuum elastic models developed in previous years were applied to gA mutants, and it was shown how substitutions in the tryptophan residue (common in nearly all membrane proteins) modulate curvature frustration. These results pave the way for quantitative modeling of the interactions of membrane proteins with the surrounding membrane. Paper 1. The Bayesian-based methodology for estimating diffusion tensors of permeants in membranes published in 2017 and extensive simulations of 7 different lipid bilayers allowed a rigorous test of the applicability of the Smoluchowski equation (equivalent, the inhomogeneous solubility-diffusion, or IHD, model) for oxygen permeation. It was shown that the Smoluchowski equation systematically underestimates first passage times (and thereby overestimates permeability) by an average of 20%. In contrast, the average lateral distance travelled in the membrane before exit from theory is nearly identical to simulation. Such assessments of the IHD model are essential for calculating reliable permeabilities of potential drugs from simulations. Paper 2.
Aim 2. Develop Simulation Methodology. Much of the work for this Aim involved force field (FF) development. Parameters of calcium-phosphate interactions for monomethyl phosphate in solution were developed with an eye to the level of clustering. The new parameters will be applied to simulations of the lipid PIP2, whose signaling function in the membrane is heavily modulated by clustering. Paper 3. FF terms necessary for the simulation of ether lipids were developed, and the results of DHPC and DPPC membrane compared. Paper 5. A rigorous and efficient method for calculation of long-range Lennard-Jones interactions was incorporated into the program CHARMM and tested on alkanes. This addition is essential for extending the CHARMM FF to polarizable systems. Paper 6. Lastly, methods for simulating lipid nanodiscs (the main focus of Aim 3) were developed and reviewed. Paper 7.
Aim 3. Simulate Complex Membranes Aim focused on simulations high density lipoprotein (HDL) models using reconstituted lipid nanodiscs (rHDL). Both natural HDL, with its scaffolding protein an apolipoprotein A1 (APOA1), and nanodiscs with APOA1 mimetic peptide were considered. Experiments of the same systems carried out by collaborators validated the simulations, and the simulations informed the interpretation of the experiments. Multi-microsecond showed how APOA1 adjusts its conformation as the nascent HDL increases in size, and disproved previous models. The symmetry breaking implies that LCAT, an essential protein in the reverse cholesterol transport pathway, has two non-equivalent binding sites. Paper 8. Simulations of the APOA1 memetic peptides showed how the ELP peptides stabilize nanodiscs by arranging in a picket-fence orientation, rather than a double belt similar to APOA1. This insight will be used to develop HDL-related therapies. Paper 4.
Showing the most recent 10 out of 52 publications
