Empirical Lipid Propensities of Amino Acid Residues in Multispan Alpha Helical Membrane Proteins Larisa Adamian, 1 Vikas Nanda, 2 William F. DeGrado, 2 * and Jie Liang 1 * 1 Department of Bioengineering, University of Illinois at Chicago, Illinois 2 Department of Biochemistry and Biophysics, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania ABSTRACT Characterizing the interactions be- tween amino acid residues and lipid molecules is important for understanding the assembly of trans- membrane helices and for studying membrane pro- tein folding. In this study we develop TMLIP (Trans- Membrane helix-LIPid), an empirically derived propensity of individual residue types to face lipid membrane based on statistical analysis of high- resolution structures of membrane proteins. Lipid accessibilities of amino acid residues within the transmembrane (TM) region of 29 structures of heli- cal membrane proteins are studied with a spherical probe of radius of 1.9 Å. Our results show that there are characteristic preferences for residues to face the headgroup region and the hydrocarbon core region of lipid membrane. Amino acid residues Lys, Arg, Trp, Phe, and Leu are often found exposed at the headgroup regions of the membrane, where they have high propensity to face phospholipid head- groups and glycerol backbones. In the hydrocarbon core region, the strongest preference for interacting with lipids is observed for Ile, Leu, Phe and Val. Small and polar amino acid residues are usually buried inside helical bundles and are strongly lipo- phobic. There is a strong correlation between vari- ous hydrophobicity scales and the propensity of a given residue to face the lipids in the hydrocarbon region of the bilayer. Our data suggest a possibly significant contribution of the lipophobic effect to the folding of membrane proteins. This study shows that membrane proteins have exceedingly apolar exteriors rather than highly polar interiors. Predic- tion of lipid-facing surfaces of boundary helices using TMLIP1 results in a 54% accuracy, which is significantly better than random (25% accuracy). We also compare performance of TMLIP with another lipid propensity scale, kPROT, and with several hydrophobicity scales using hydrophobic moment analysis. Proteins 2005;59:496 –509. © 2005 Wiley-Liss, Inc. Key words: membrane protein; accessible surface; lipid propensity; alpha shape; hydropho- bicity INTRODUCTION The folding of helical membrane proteins involves the burial of residues from transmembrane (TM) helices in a lipid bilayer and the assembly of TM helices. 1–3 The contribution from side chains interacting with their envi- ronment reflects the energetic cost or gain due to the exposure of the residue to the lipid bilayer, or to the burial of the residue within the protein core. The contribution from interhelical interactions reflects the energetic cost or gain of various types of two-body and many-body interac- tions between transmembrane helices. The entropic effects include, among other terms, the restriction of the conforma- tions of connected backbones and side chains. Quantitative estimation of these contributions is essential for model studies of membrane protein folding. Here we estimate the free-energy cost or gain associated with the burial or exposure of different amino acid residue types to the lipid bilayer environment. This is an impor- tant endeavor, both for understanding the features stabi- lizing membrane proteins as well as the prediction of membrane protein structures. For example, an energetic scale for lipid exposure could be useful in differentiating properly folded from mis-folded structures generated by either ab initio or threading approaches to membrane protein structure prediction. Different lipid-contacting re- gions of a TM helix face either the highly hydrophobic hydrocarbon core or the more polar headgroup region of the lipid bilayer. Thus, different types of amino acid residues are likely to have different propensities for expo- sure to lipids at distinct regions of the helix–lipid inter- faces. Indeed, Spencer and Rees 4 showed that helical TM proteins exhibit a central 20 Å-wide region with greater than 90% of its surface area contributed by carbon atoms, and very few formally charged atoms. On either side of this region the polarity of the protein increases in an approxi- mately linear manner, reaching the distribution observed in water-soluble proteins after another 10 Å has been traversed. Grant sponsor: the National Science Foundation; Grant numbers: CAREER DBI0133856 and DBI0078270; Grant sponsor: the National Institute of Health; Grant numbers: GM68958, HL07971-0, and GM60610. *Correspondence to: Jie Liang, Department of Bioengineering, University of Illinois at Chicago, M/C563, 835 S. Wolcott, Chicago, Illinois 60612-7340. E-mail: jliang@uic.edu; and William DeGrado, Department of Biochemistry and Biophysics, School of Medicine, University of Pennsylvania, 3700 Hamilton Walk, Philadelphia, Penn- sylvania 19104. E-mail: wdegrado@mail.med.upenn.edu. Received 30 July 2004; Accepted 22 December 2004 Published online 23 March 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/prot.20456 PROTEINS: Structure, Function, and Bioinformatics 59:496 –509 (2005) © 2005 WILEY-LISS, INC.