Feigenson Lab

MD Simulations

Trends in Phase Separation for CG Molecular Dynamics of Four Component Lipid Bilayers

Our lab is particularly interested in studying bilayers containing 4 lipid components. Here we present Molecular Dynamics simulation results of lipid bilayers modeled using the Martin Coarse Grained forcefield. These bilayers contain 4,608 lipids, including 18:0,18:0-PC (red)/[18:0,18:2-PC (green)/18:2,18:2-PC (blue)]/Cholesterol (gray) in a ratio of 40/40/20.

The amount of two components - the high melting lipid (18:0,18:0-PC) and cholesterol - are kept fixed, and the only paramater varied from one bilayer to the next is the relative amounts of two different low melting lipids: 18:0,18:2-PC (does not favor large domains when mixed with 18:0,18:0-PC and Cholesterol) and 18:2,18:2-PC (favors large domains when mixed with 18:0,18:0-PC and Cholesterol). The relative amount of all low melting lipid that is 18:2,18:2-PC is defined as rho. We are interested in the transition from no domains to macroscopic phase separation as rho is increased from 0 to 1. For the simulation videos in this channel, the fixed composition for the mixtures of 18:0,18:0-PC/[18:0,18:2-PC/18:2,18:2-PC]­/Chol is 40/40/20.

The simulation results are for 15 microseconds of simulation run time. Each individual bilayer has a different rho value, as can be seen in the different amounts of 18:0,18:2-PC (green) and 18:2,18:2-PC (blue) they each contain. Rho=0 (no 18:2,18:2-PC) is in the top left, rho=1 (no 18:0,18:0-PC) is in the bottom right. Rho increases by 0.1 from one simulation to the next, going from left to right.

As can be seen in these results, even small amounts of 18:0,18:2-PC, which does not favor domain formation, cause domains to break up (see for instance rho=0.8). We also see that lipid clustering increases with increasing rho, and at about rho=0.6, very large stable domains are present.

We are currently studying what is changing at a molecular level at the interface between these large domains as rho is varied.



Phase Separation in Lipid Bilayers: Non-ideal Mixing in CG Molecular Dynamics

Here we present Molecular Dynamics simulation results of a lipid bilayer modeled using the Martin Coarse Grained forcefield. This bilayer contains 4,608 lipids, including 18:0,18:0-PC (red), 18:0,18:2-PC (green) and Cholesterol (gray) in a ratio of 40/40/20. The video shows the two leaflets of the bilayer over 15 microseconds of simulation time. These lipids do not phase separate, but rather exhibit non-ideal mixing where large clumps are visible, but transient.

We are particularly interested in studying bilayers containing 4 lipid components.The amount of two components - the high melting lipid (18:0,18:0-PC) and cholesterol - are kept fixed, and the only paramater varied is the relative amounts of two different low melting lipids: 18:0,18:2-PC (does not favor large domains when mixed with 18:0,18:0-PC and Cholesterol) and 18:2,18:2-PC (favors large domains when mixed with 18:0,18:0-PC and Cholesterol). The relative amount of all low melting lipid that is 18:2,18:2-PC is defined as rho. We are interested in the transition from no domains to macroscopic phase separation as rho is increased from 0 to 1. For the simulation videos in this channel, the fixed composition for the mixtures of 18:0,18:0-PC/[18:0,18:2-PC/18:2,18:2-PC]­/Chol is 40/40/20.

This simulation, containing no 18:2,18:2-PC, is therefore an example of a rho=0 bilayer.



Phase Separation in Lipid Bilayers: The Onset of Phase Separation in CG Molecular Dynamics

Here we present Molecular Dynamics simulation results of a lipid bilayer modeled using the Martin Coarse Grained forcefield. This bilayer contains 4,608 lipids, including 18:0,18:0-PC (red), 18:0,18:2-PC (green), 18:2,18:2-PC (blue) and Cholesterol (gray) in a ratio of 40/20/20/20. The video shows the two leaflets of the bilayer over 15 microseconds of simulation time. These lipids are beginning to phase separate over the simulation time, as can be seen in the large stable clusters of lipids. Though the leaflets are not very correlated, they both do show larger clusters compared to simulations that lack 18:2,18:2-PC (rho=0; see below).

We are particularly interested in studying bilayers containing 4 lipid components.The amount of two components - the high melting lipid (18:0,18:0-PC) and cholesterol - are kept fixed, and the only paramater varied is the relative amounts of two different low melting lipids: 18:0,18:2-PC (does not favor large domains when mixed with 18:0,18:0-PC and Cholesterol) and 18:2,18:2-PC (favors large domains when mixed with 18:0,18:0-PC and Cholesterol). The relative amount of all low melting lipid that is 18:2,18:2-PC is defined as rho. We are interested in the transition from no domains to macroscopic phase separation as rho is increased from 0 to 1. For the simulation videos in this channel, the fixed composition for the mixtures of 18:0,18:0-PC/[18:0,18:2-PC/18:2,18:2-PC]­/Chol is 40/40/20.

This simulation, containing equal amounts of 18:0,18:2-PC and 18:2,18:2-PC, is therefore an example of a rho=0.5 bilayer.



Phase Separation in Lipid Bilayers: Macroscopic Phase Separation in CG Molecular Dynamics

Here we present Molecular Dynamics simulation results of a lipid bilayer modeled using the Martin Coarse Grained forcefield. This bilayer contains 4,608 lipids, including 18:0,18:0-PC (red), 18:2,18:2-PC (blue) and Cholesterol (gray) in a ratio of 40/40/20. The video shows the two leaflets of the bilayer over 15 microseconds of simulation time. These lipids completely phase separate over the course of the simulation, and the two leaflets are correlated.

We are particularly interested in studying bilayers containing 4 lipid components.The amount of two components - the high melting lipid (18:0,18:0-PC) and cholesterol - are kept fixed, and the only paramater varied is the relative amounts of two different low melting lipids: 18:0,18:2-PC (does not favor large domains when mixed with 18:0,18:0-PC and Cholesterol) and 18:2,18:2-PC (favors large domains when mixed with 18:0,18:0-PC and Cholesterol). The relative amount of all low melting lipid that is 18:2,18:2-PC is defined as rho. We are interested in the transition from no domains to macroscopic phase separation as rho is increased from 0 to 1. For the simulation videos in this channel, the fixed composition for the mixtures of 18:0,18:0-PC/[18:0,18:2-PC/18:2,18:2-PC]­/Chol is 40/40/20.

This simulation, containing no 18:0,18:2-PC, is therefore an example of a rho=1 bilayer.