Properties of integral membrane protein structures: Derivation of an implicit membrane potential

Distributions of each amino acid in the trans‐membrane domain were calculated as a function of the membrane normal using all currently available α‐helical membrane protein structures with resolutions better than 4 Å. The results were compared with previous sequence‐ and structure‐based analyses. Calculation of the average hydrophobicity along the membrane normal demonstrated that the protein surface in the membrane domain is in fact much more hydrophobic than the protein core. While hydrophobic residues dominate the membrane domain, the interfacial regions of membrane proteins were found to be abundant in the small residues glycine, alanine, and serine, consistent with previous studies on membrane protein packing. Charged residues displayed nonsymmetric distributions with a preference for the intracellular interface. This effect was more prominent for Arg and Lys resulting in a direct confirmation of the positive inside rule. Potentials of mean force along the membrane normal were derived for each amino acid by fitting Gaussian functions to the residue distributions. The individual potentials agree well with experimental and theoretical considerations. The resulting implicit membrane potential was tested on various membrane proteins as well as single trans‐membrane α‐helices. All membrane proteins were found to be at an energy minimum when correctly inserted into the membrane. For α‐helices both interfacial (i.e. surface bound) and inserted configurations were found to correspond to energy minima. The results demonstrate that the use of trans‐membrane amino acid distributions to derive an implicit membrane representation yields meaningful residue potentials. Proteins 2005. © 2005 Wiley‐Liss, Inc.