The Protein that Held Back the Dye: Annexin's Effect on Membrane Permeability

Abstract

Using Bacillus thuringiensis phosphatidylinositol-specific phospholipase C (BtPI-PLC) as a model amphitropic protein, we are investigating how membrane structure and composition affect protein-membrane interactions. Previous work showed that BtPI-PLC specifically binds to phosphatidylcholine (PC)-rich membranes and preferentially interacts with unilamellar vesicles with high curvature. In this work, we monitored single fluorescently labeled BtPI-PLC proteins as they cycled on and off surface-tethered phosphatidylglycerol (PG)/PC small unilamellar vesicles (SUVs) using total internal reflection fluorescence (TIRF) microscopy. The residence times on vesicles along with vesicle size information, based on vesicle fluorescence intensity, reveal the time scales of protein-membrane interactions as well as the curvature dependence. BtPI-PLC residence times on SUVs average 300 ms, similar to published residence times (300-400 ms) for other amphitropic proteins that transiently interact with cell surfaces. The kinetics of PI-PLC/membrane interactions is well explained by a simple two state binding model with dissociation and association rate constants averaging 3 s À1 and 0.6 mM À1 s À1 respectively. In addition fluorescence correlation spectroscopy (FCS) measurements indicate that introducing lipid packing defects PG/PC SUVs by incorporating low mole percentages of dioleoylglycerol (DOG) enhances BtPI-PLC binding to SUVs. By combining these single molecule fluorescence results with previous biophysical measurements and molecular dynamics simulations, we have developed a quantitative model showing how the bacterial virulence factor Bt-PI-PLC interacts with cell membranes in molecular detail.