Methods for the Design and Analysis of Sedimentation Velocity and Sedimentation Equilibrium Experiments with Proteins

Data quality comparison between absorbance (A, C, E) and intensity (B, D, F) data: A sedimentation experiment of a solution containing 0.1 OD ovalbumin was scanned at 280 nm in absorbance mode (panel A) and fitted by two-dimensional spectrum analysis with time- and radially invariant noise removal. The residuals of this fit are shown in panel C, and the noise corrected data are shown in panel E. After the experiment, the solution was shaken up and rescanned in intensity mode under identical run conditions, producing corresponding experimental scans in panel B, residuals in panel D and noise-subtracted data in panel F. The increased noise that can be seen in the absorbance experiment is due to the convolution of two stochastic measurements (reference and sample), which amplifies the stochastic noise approximately by a factor of Ö(2). In this case, the RMSD of the absorbance experiment was 3.5543 × 10^–3 and the RMSD of the intensity experiment was 2.3705 × 10^–3. The additional noise seen in panel B arises from time-invariant noise components contributed by optical components other than cell windows, which are subtracted out in the absorbance experiment. Regions in the residuals where deviations exceed the average are due to narrow, high-amplitude time invariant noise signals (most likely scratches on the cell windows) that are poorly reproduced by the scanning optics due to lack of precision in the radial scanning system.

Global extinction fit (thin black lines) of 15 wavelength scans (gray circles) at five different loading concentrations. Each loading concentration is scanned three times with 1-nm increments and fitted globally to the same sum of Gaussian terms to represent the intrinsic extinction profile with the UltraScan “Global Extinction Fit” module. Each loading concentration is represented by the same Gaussian sum, but with a different amplitude. The intrinsic extinction profile (heavy black line) of the protein is normalized by a known extinction coefficient (typically at 280 nm).

van Holde-Weischet integral distribution plots for the ovalbumin sample shown in Figure 7.13.3, measured at three different loading concentrations. Filled squares represent 0.1 OD at 280 nm, open circles represent 0.3 OD at 280 nm, and filled triangles represent 0.9 OD at 280 nm. The vertical shape of the s-value distribution indicates homogeneity. Identical results for all three concentrations indicate absence of reversible self-association in this concentration range (in contrast to the data shown in Fig. 7.13.8).