IDENTIFYING NONPOLAR TRANSBILAYER HELICES IN AMINO ACID SEQUENCES OF MEMBRANE PROTEINS
D. M. Engelman, T. A. Steitz, and A. Goldman
11/06/2014
Annual review of biophysics and biophysical chemistry
In recent years amino acid sequences for many integral membrane proteins have been determined. At the same time theoretical arguments and experimental evidence have accumulated to indicate that trans bilayer helices are a major motif in integral membrane protein structure. It may be possible to determine the location of such transmembrane helices directly from amino acid sequences using scales of polarity and sequential search protocols. The purpose of this review is to examine the approaches that have been used and to assess their utility. It is an opportune time to consider the issues, since the structure of three integral membrane proteins that are contained in the photosynthetic reaction center of Rhodopseudomonas viridis have been determined at high resolution. This new structural information, combined with increasing evidence concerning the structure of bacteriorhodopsin, permits a critical test of the main ideas involved in searching sequences for trans bilayer structural elements. We find that search procedures based on a moving window that scans a sequence twenty residues at a time are suitable for finding the transbilayer helices that are known to exist. For very nonpolar helices separated by polar polypeptides, many of the proposed polarity scales succeed equally well; where the helices contain more polar groups, however, the choice of scales becomes critically important. In our discussion we consider the arguments that support the notion that helical structure will be a dominant motif in integral membrane protein organization. We introduce and discuss the problem of suitable scaling of amino acids in terms of their polar and nonpolar characteristics, and discuss further the use of such scales in prediction of protein structure. Finally, we examine the cases in which the validity of predictions can be assessed. It is our contention that a suitable scale and protocol can lead to the successful identification of transmembrane helical structures in integral membrane proteins.