Effect of spectral window size on circular dichroism spectra deconvolution of proteins

Abstract

The recently developed Convex Constraint Algorithm (CCA), which has been found to be efficient for CD spectra deconvolution of proteins, was used to examine the importance of the spectral window size and its effects on the deconvolution sensitivity. Using the “largest” protein CD spectra data base ever published (A. Toumadje, S.W. Alcorn and W.C. Johnson, Jr., AnaL Biochem. 200 (1992) 321-331) a systematic “spectratruncation” was performed. The reduced spectral data sets were deconvoluted and the Pearson product-moment correlation (r), as well as the RMS values, were calculated. It was found that the spectral window size enlargement below 180 nm has only minor secondary effect on the accuracy of the deconvolution, since the r and RMS values were nearly insensitive to the broadening of the spectral window. By contrast, these values monitored the decrease of the deconvolution capacity for the analysis when the 180-200 nm spectral region was likewise eliminated. This observation harmonizes perfectly with the recent observation of S.Y. Venyaminov, I.A. Baikalov, C.-S.C. Wu and J.T. Yang (AnaL Biochem. 198 (1991) 250-255) and with the pioneering observation of N. Greenfield and G.D. Fasman (Biochemistry 8 (1969) 4108-4116), namely, by taking only the CD spectral region above 200 nm, the α-helix content may be estimated with an acceptable accuracy. It is therefore concluded that the development and the analysis of a large protein data base (e.g. dozens of spectra), incorporating as many “dissimilar proteins” as possible, may result in a better understanding of the relationship between the CD spectra and the secondary structural elements of proteins.