Use of Synthetic Peptides for the Study of Membrane Protein Structure

Abstract

Living cells maintain tenfold or greater concentration gradients of Na +, K⁺, and Ca²⁺ across an ion-impermeant hydrocarbon layer approximately 30 Å thick in their lipid bilayer membranes. Protein “pumps” embedded in these membranes create the gradients, and protein ion channels modulate their discharge in response to various stimuli. Physiological processes involving ion channels range from the very elementary sensory systems of single-celled organisms such as Escherichia coli (e.g., Saimi et al., 1988) to the massively interconnected information-processing network of the mammalian brain (e.g., Nicoll, 1988). Historically, ion channels were proposed to be the elementary ion conduction elements in nerve impulse transmission (reviewed by Hille, 1984). Decades of study of nervous system and muscle physiology have since identified hundreds of functionally distinct ion channels. More recently, recombinant DNA technology has provided the amino acid sequences of many ion channel proteins. They appear to form a small number of “families” (Numa, 1989) characterized both by function and by associated, distinguishing features of their sequences. The apparent correlation of sequence features with channel functional type has prompted much curiosity and speculation concerning the three-dimensional structure and mechanisms of functioning of ion channel proteins (see, e.g., Montal, 1990). Most of these proteins, however, contain thousands of amino acid residues. Given our current level of understanding of protein folding, such proteins are far too large and complicated for molecular modeling to provide a satisfactory level of structural detail. With three-dimensional crystal structures, of course, one might hope to pinpoint the features relevant to channel function, but crystals of ion channel proteins have, so far, not provided sufficiently high-resolution diffraction information to resolve sequence-related details. Consequently, other ways are needed to obtain structure—function information from sequence data. Comparative studies of natural sequence variations within channel families (e.g., Butler et al., 1989), functional analysis of site-directed mutations (e.g., Imoto et al., 1988; Stühmer et al., 1989), and studies of synthetic peptides with sequences based on specific channel proteins (e.g., Oiki et al., 1988) are all being used.