UBC Theses and Dissertations
A study of secondary structure predictive methods for proteins and the relationship between physical-chemical properties and enzymatic activity of some aspartyl proteinases Yada, Rickey Yoshio
In the first of two studies, four algorithms were assessed for their ability to accurately determine protein secondary structure from circular dichroism spectral data in relation to X-ray data. The majority of the algorithms examined showed highly significant (P<0.001) correlation coefficients for ∝-helix and β-sheet determination while the correlation coefficients using the simplex optimization algorithm with n = variable (in the presence of concanavalin A) were significant at P<0.05. None of the algorithms examined showed significant (P>0.05) correlation coefficients for β-turn determination. Significant (P<0.05) correlation coefficients were obtained for random coil determination from both the simplex optimization algorithm (in the absence of concanavalin A) and the algorithm of Chang et al. (1978). The simplex optimization algorithm with n = 10.4 compared favourably to the method of Chang et al. (1978) in the absence of concanavalin A, and may serve as an alternative algorithm to solve least squares. The algorithm of Provencher and Glockner (1981) demonstrated the greatest versatility for secondary structure determination of the four algorithms examined. In the second study, the relationship between physical-chemical properties and the enzymatic activity of aspartyl proteinases was investigated. Using the diagonal plot method, pepsin and chymosin showed the highest degree of primary sequence homology while active site regions were highly homologous between the aspartyl proteinases examined. Secondary structure prediction methods indicated that chymosin, pepsin, peni-cillopepsin and Mucor miehei proteinase had relatively high proportions of β-sheet, with active site aspartic acid residues located in β-turn regions. Near-UV and far-UV CD spectral analysis indicated that changes in spectra occurred in the neutral to alkaline pH range. Secondary structure determination from far-UV CD spectral data demonstrated that the β-sheet fraction of the aspartyl proteinases decreased above pH 6.3. The milk-clotting to proteolytic activity ratio of the aspartyl proteinases decreased with increasing pH. Principal components derived from various structural and intrinsic parameters allowed for the classification of aspartyl and non-aspartyl proteinases. Regression of the milk-clotting to the proteolytic activity ratio on the principal components indicated that a high ratio may be partially dependent on relatively high hydrophobicity, β-sheet and low charge.
Item Citations and Data