A model, based on amino acid sequence data, is proposed for the organization of the myelin proteolipid in myelin membrane. The model has three distinctive features: three trans-membrane segments that traverse the lipid bilayer, two cis-membrane domains that enter and exit the same side of the membrane, and a highly charged segment resembling myelin basic protein on the cytoplasmic side of the membrane. It is proposed that the cis-membrane domain(s) can promote the formation and stabilization of the multilamellar myelin structure by hydrophobic interaction with the apposite bilayer across the extracellular space.In the central nervous system, myelin is formed as an extended, modified oligodendroglial plasma membrane that spirals around the axon to form a multilamellar structure (1). During maturation, the membranes become compacted with a close apposition of both the internal and external faces of the membrane. The relative thicknesses of the bilayer, cytoplasmic, and extracellular spaces are 47 A, 30 A, and 30 A, respectively (2,3).Although the myelin sheath is characterized by a relatively high proportion of lipid (70-80%), it is the myelin proteins that provide specificity and perform an important, albeit illdefined role in the maintenance of myelin structure. The predominant proteins are the water-soluble extrinsic myelin basic protein (Mr 18,000), characterized by a high proportion of basic and other polar amino acids (4,5), and the hydrophobic, chloroform/methanol-soluble proteolipid (Mr 30,000) (4, 6). The basic protein is well-characterized (4, 5); however, the proteolipid has resisted structure elucidation, and only recently has its amino acid sequence been determined (7-9). In this article we propose a model ( Fig. 1) for bovine brain myelin proteolipid, based on amino sequence data and analogy with other proteins, and suggest how the proteolipid may help to stabilize the myelin sheath.
RESULTS AND DISCUSSIONGeneral Structural Features. The most notable characteristic of the proteolipid sequence (Fig. 2) is the clustering of hydrophobic amino acids into distinct domains. A plot of amino acid hydropathy (11) (Fig. 3) shows four hydrophobic domains of about 30 amino acids each, which alternate with polar segments containing most of the charged and neutral hydrophilic residues. The polar domains also contain all of the predicted ,B-turns (12), suggesting that these regions are extensively folded. These features suggest a model, such as has been proposed for bacteriorhodopsin (13) and other intrinsic membrane proteins (14), wherein the polypeptide chain passes repeatedly through the lipid bilayer, the hydrophobic regions being embedded in the lipid and the polar domains exposed on the external and internal faces of the membrane (Fig. 1). For purposes of discussion, various domains in the proteolipid are defined in terms of polarity, proposed orientation in the membrane, and sequence homology. Thus, Ti (residues 59-90), T2 (residues 151-177) and T3 (residues 238-267) are hydrophobic trans-membrane se...