PREDICTION OF THE SECONDARY STRUCTURE OF MYELIN BASIC-PROTEIN

The probable structure of the myelin basic protein [isolated from human, bovine, rat, rabbit, guinea pig, monkey, mouse and chicken] was studied by the application of 3 procedures currently in use to predict the secondary structures of proteins from knowledge of their amino acid sequences. To increase the accuracy of the predictions, the amino acid substitutions that occur in the basic protein from different species were incorporated into the predictive algorithms. It was possible to locate regions of probable .alpha.-helix, .beta.-structure, .beta.-turn and unordered conformation (coil) in the protein. One of the predictive methods introduces a bias into the algorithm to maximize or minimize the amounts of .alpha.-helix and/or .beta.-structure present; this made it possible to assess how conditions such as pH and protein concentration or the presence of anionic amphiphilic molecules could influence the protein''s secondary structure. The predictions made by the 3 methods were in reasonably good agreement with one another. They were consistent with experimental data, provided that the stabilizing or destabilizing effects of the environment were taken into account. According to the predictions, the extent of possible .alpha.-helix and .beta.-structure formation in the protein is severely restricted by the low frequency and extensive scattering of hydrophobic residues, along with a high frequency and extensive scattering of residues that favor the formation of .beta.-turns and coils. Neither prolyl residues nor cationic residues per se are responsible for the low content of .alpha.-helix predicted in the protein. The principal ordered conformation predicted is the .beta.-turn. Many of the predicted .beta.-turns overlap extensively, involving in some cases up to 10 residues. In some of these structures it is possible for the peptide backbone to oscillate in a sinusoidal manner, generating a flat, pleated sheetlike structure. Cationic residues located in these structures would appear to be ideally oriented for interaction with lipid phosphate groups located at the cytoplasmic surface of the myelin membrane. An analysis of possible and probable conformations that the triproline sequence could assume questions the popular notion that this sequence produces a hairpin turn in the basic protein.