Membrane-targeting proteins are recruited to specific membrane surfaces during cell signaling events. These interactions frequently involve the recruitment of multiple lipids including phosphatidylserine (PS), phosphatidylinositol-4,5-bisphosphate (PIP2), phosphatidylinositol-3,4,5-trisphosphate (PIP3) and diacylglycerol (DAG). Interactions with proteins are highly regulated. The pleckstrin homology (PH) domain specifically binds PIP3, and several membrane-targeting proteins have multiple PH domains. As a more complex example, protein kinase C has three specific binding sites, one for each of PS, PIP3 and DAG, and all three are required for full activation. Determining the number and types of bound lipids is thus crucial to understanding the activity of membrane-targeting proteins.
I have developed computational and theoretical models to determine a precise and readily measurable relationship between the lateral diffusion constant of a protein and the number of specifically bound lipids[1,2]. The correspondence with experimental work done in collaboration with Joe Falke and Jeff Knight is excellent. The computational models involve the Martini force field, and are simulations are run in GROMACS and CHARMM. The analytical results are based primarily on the Saffman-Delbruck model of a cell membrane, using the formalism of velocity response functions.
This work has been highlighted in the Biophysical Society newsletter, and was the second most read Biophysical Journal paper in all of July-December 2010.
I am working to extend the model to include larger systems as well as to give insight into specific vs. non-specific binding. I have also performed an extensive set of coarse-grained and all-atom simulations to investigate the general physical principles that govern membrane diffusion.