R o

1720 1700 1680 1660 1640 1620

1600

WAVENUMBER, cnr>

Figure 7.6 Best fit by nonlinear regression analysis of the Fourier deconvoluted lysozyme spectrum using a resolution enhancement factor (REF) of 3.8. Outer envelope lines as in Fig. 7.5. (Reprinted with permission from [9], copyright by the American Chemical Society.)

WAVENUMBER, cm1

Figure 7.7 Best fit by nonlinear regression analysis of the Fourier deconvoluted lysozyme spectra using a resolution enhancement factor of 2.5. Outer envelope lines as in Fig. 7.5. (Reprinted with permission from [9], copyright by the American Chemical Society.)

WAVENUMBER, cm1

Figure 7.7 Best fit by nonlinear regression analysis of the Fourier deconvoluted lysozyme spectra using a resolution enhancement factor of 2.5. Outer envelope lines as in Fig. 7.5. (Reprinted with permission from [9], copyright by the American Chemical Society.)

B.3. Structural Comparisons with Crystal Structures of Proteins

A major advantage of FT-IR analyses for assessments of the secondary structure of proteins is that they can be done using only small amounts of protein in solutions [9,10], in films [12], and in membranelike environments [12], The classical method of determining structures of proteins is X-ray crystallography, which obviously requires a good protein crystal. In Table 7.4, we saw that the FT-IR analysis of the secondary structure of lysozyme gives a good correspondence with the X-ray structure. Further validation of the FT-IR method has been achieved by comparison of the FT-IR analyses of 14 proteins with their known crystal structures [10].

Computing fractions of structural features from X-ray crystal structures is somewhat imprecise [10]. Not only are fractions of a helix, turn, and extended conformations important, but the lengths of the helical and extended features as well as the presence of internal backbone hydrogen bonding may also contribute to the component bands of the amide I region. Thus, exact comparisons between FT-IR and X-ray crystallographic results will be subject to error. Also, the result may differ because of real differences between crystal structures and structures of the same protein in aqueous solutions. In other words, the X-ray results do not constitute flawless primary standards of comparison for the FT-IR results, but they are the best we have at present.

In the discussion that follows, the traditional Ramachandran plot calcu lated from the X-ray crystal structure was used to estimate the fractions of structural features from data in the Brookhaven Protein Data Bank. Details can be found in [9] and [10].

FT-IR and X-ray crystallographic comparisons for the 14 proteins and polypeptides are summarized in Figures 7.8 to 7.11. Slopes of 1.0 and correlation coefficients (r) of 1.0 in these plots would indicate perfect correlations. The FT-IR and X-ray results for the percent of helix (Figure 7.8) are well correlated, with a slope of 0.91 and r close to 1. The scatter in this plot increases below 20% helix. The correlations for the percent of turns (Figure 7.9) and percent of extended conformations (Figure 7.10) are even better, with r values of 0.99 and slopes close to 1. The correlation for the percent disordered (Figure 7.11), is not as good as for the other structural

Figure 7.8 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of helical conformation. Proteins and % helix from FT-IR are:

Figure 7.8 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of helical conformation. Proteins and % helix from FT-IR are:

20 40 60 80 100

% helix from X-ray data

20 40 60 80 100

% helix from X-ray data

Protein

% helix

Hemoglobin

Myoglobin

Cytochrome C

Lysozyme

Ribonuclease

Papain

P-tripsin inhibitor a-Chymotrypsin Trypsin Elastase

Carbonic Anhydrase /3-Lactoglobulin Conconavalin A Oxytocin

15.8

14.1

14.2

Figure 7.9 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of extended conformation. Proteins are the same as in Figure 7.8.

Figure 7.9 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of extended conformation. Proteins are the same as in Figure 7.8.

features, with a smaller slope and r value. However, recall that the disordered features are not directly estimated by the X-ray analysis, so that error in both the X-ray and FT-IR data contribute to the scatter.

These results show overall good agreement of the FT-IR analysis with the X-ray crystal structures. Analyses of 14 proteins whose crystal structures are known showed agreement of secondary structure to within about 5% between the experimental FTIR and X-ray estimations. These correlations provide confidence in the reliability of the FT-IR secondary structural

% turns from X-ray data

Figure 7.10 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of turns and twists. Proteins are the same as in Figure 7.8.

% turns from X-ray data

Figure 7.10 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of turns and twists. Proteins are the same as in Figure 7.8.

Figure 7.11 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of disordered conformation. Proteins are the same as in Figure 7.8.

% disordered X-ray data

Figure 7.11 Correlation of results of FT-IR secondary structural analysis with X-ray crystal structure for the percent of disordered conformation. Proteins are the same as in Figure 7.8.

analysis. They suggest that such an analysis can provide an approximate picture of protein conformation in solutions. Large conformational changes, e.g., partial denaturation, can be readily identified by this type of analysis.

0 0

Post a comment