Prediction of the secondary structure of Bowman-Birk soybean protease inhibitor from LTS sequence

The secondary structure of Bowman-Birk soybean inhibitor (BBI) has been predicted from its amino acid sequence (71 residues) using the statistical method of Chou and Fasman (1974) as well as the informational method of Garnier et al. (1978). According to both methods, no alpha-helical region is predicted in BBI. Short beta-strands at 11-15, 39-43, 48-52 and 57-59 are predicted by Chou and Fasman's method, and at 11-14, 39-43, 48-52 and 57-59 by the method of Garnier et al. Predicted beta-turn tetrapeptides are residues at 1-4, 6-9, 14-17, 18-21, 24-27, 30-33, 36-39, 60-63, 63-66 and 68-71, according to the method of Chou and Fasman. These predictive results indicate that BBI consists essentially of beta-turn, beta-strand, and unordered structures, which represent about 56, 25 and 19% (by difference) of its secondary structure, respectively. The predicted reverse turn amount in BBI is very high, in comparison with that found usually in proteins. No fitting of the experimental circular dichroism spectrum (in the 195-240 nm region), which was reported by Kay (1976), has proved satisfactory when using the parameters for helical, beta-sheet, beta-turn and unordered conformations proposed by several authors.