The G-protein coupled receptors (GPCRs) are an enormous superfamily of transmembrane signaling proteins, comprising some 800 members. GPCRs are found in every imaginable physiological context, from vision, to immune response, to taste and smell, to the nervous system. It is estimated that the GPCRs account for the targets of nearly half of the drugs presently on the market, making them perhaps the most pharmacologically interesting protein family.

In collaboration with several experimental groups (Ilya and Kandice Levental, UVA; Anne Robinson, CMU; Noah Malmstadt, USC) we are working to understand the mechanism behind lipid-dependent GPCR function. Our efforts are grounded in a modern view of the lipidomic complexity of mammalian membranes. Our long term goal is to exploit variation in lipidomes for therapeutic ends.

Our membranes are complex, asymmetric mixtures of lipids, dense with protein. Historically, biophysicists have relied on simple bilayer models, and then hoped that what we learn in this context transfers to the complexity of real membranes. As experimental techniques have advanced, so too must our simulations. When does lipidomic complexity matter, and when are simpler model systems good enough?

Our group has pioneered the all-atom simulation of complex lipid bilayers, using a series of allocations on the Anton special purpose machine. Anton is ideal for these simulations, as the typical mixing times are ten microseconds and longer. A key focus has been on the comparison of experimental and simulated observables, including 2H NMR spectra, small-angle neutron scattering, polarity sensitive dyes such as Laurdan, and small molecule permeabilities.