| Abstract: | We have used mean hydrophobicity and hydrophobic moment calculations to predict the receptor binding domains in apolipoprotein E and in the low density lipoprotein apolipoprotein B-E receptor. In apolipoprotein E, two receptor binding domains, residues 136-160 and 214-236, having a high hydrophilicity and a high mean helical hydrophobic moment, were identified. The first domain has been located experimentally and mutations influencing the hydrophobicity parameters of the binding site have been shown to affect the receptor binding. The second domain is probably, either separately or in combination with the first domain, involved in receptor binding or in heparin binding. In the low density lipoprotein apolipoprotein B-E receptor, six protein domains were identified. In the first domain (residues 1-371), eight hydrophilic maxima, organized in pairs through disulfide bonds, form the four experimentally observed receptor binding sites. These sites consist of repeats of 26 amino acids but differ from those reported by others [Yamamoto, T., Davis, C. G., Brown, M. S., Schneider, W. J., Casey, M. L., Goldstein, J. W. & Russell, D. W. (1984) Cell 39, 27-38]. The second, more hydrophobic, domain (residues 372-640) forms the core of the receptor, explaining its homology with the precursor of mouse epidermal growth factor, while the cysteine residues in the third domain (residues 641-699), interacting with those of the first domain, further stabilize the molecule. Beyond the fourth hydrophilic domain (residues 700-767), to which carbohydrates are linked, a very hydrophobic membrane spanning region (residues 768-789) could be detected easily. The last domain (residues 790-839), situated in the cytoplasma, contains hydrophilic maxima, as this region might interact with clathrin-related proteins. These data suggest that hydrophobicity analysis can detect and predict protein domains: hydrophilic receptor sites as well as hydrophobic core-forming and membrane-spanning regions. |
