Measuring Side-Chain Hydrophobicity using a Whole Transmembrane Protein in Lipid Bilayers

Abstract

Membrane protein behavior should be predictable from first principles of physical chemistry. One such first principle is the hydrophobic effect, which may be the foremost reason membrane proteins are stable in lipid bilayers. Several methods exist for estimating the hydrophobicity of proteins in membranes by modeling the energies felt by individual amino acid side-chains in hydrophobic environments. However, none of these existing methods have measured side-chain energetics directly from a whole multi-pass membrane protein in a lipid bilayer. Consequently, the different hydrophobicity scales that have been proposed to be relevant for membrane proteins disagree on the magnitude of the first principles, especially the principles controlling the interaction of ionizable side-chains (such as Arginine) with membranes. Here, we establish a new whole protein hydrophobicity scale, which we measured using rigorous equilibrium thermodynamic methods with a multi-pass transmembrane protein that was folded and inserted in a phospholipid bilayer. Our protein was the Outer Membrane Phospholipase A (OMPLA) from E. coli . This beta-barrel protein yields results that are relevant for all membrane proteins, even alpha-helical ones, because the contribution of the backbone conformation is subtracted from our final measurements. We were also able to use OMPLA to measure how the energetics of side-chain insertion vary with depth in the bilayer. Both Arginine and Leucine have their most extreme insertion energies when they are closest to the middle of the hydrophobic region of OMPLA. Further, we carried out a double mutant cycle with Arginines and discovered that the insertion of a second Arginine is aided by the insertion of the first. This result is particularly relevant for understanding the function of the voltage sensing domains of some ion channels, which may involve multiple Arginines penetrating the lipid bilayer.