Probing protein conformation by infrared spectroscopy.

Infrared Spectroscopy has become a usual technique in studies of protein conformation. The use of modern instruments based on interferometric devices together with the application of Fourier-transform techniques makes possible to obtain high signal-to-noise ratio spectra. These advantages allow to apply resolution enhancement techniques such as deconvolution or derivation in order to derive information essential for the quantitative analysis of the spectra 111. Amide I band, composed mainly by peptidic C=O stretching vibrations and located between 1700 and 1600 cm-I, is the most used band in quantitative studies. In principle, once the number and position of the amide I band are known with the resolution enhancement techniques, the component bands are obtained through a least squares iterative procedure. However, deconvolved spectra rather than the original ones are usually employed in the fitting procedure since the shapeless amide I band is not pure but a mixture of gaussian and lorentzian !or Cauchy), whereas the deconvolved spectrum is in principle gaussian, being the fitting is easier. Moreover, even if the spectra in H20 can be obtained, the use of small pathlengths (6-10 pm) makes easier to acquire the spectra in D,O where a better signal-to-noise ratio is obtained. In the present work, we have compared the results obtained by fitting a spectra using the original or the deconvolved trace, and we have also established that if only spectra in D,O media are used, the assignment of the band components to protein structures can be easily misinterpreted. The amide I band of a spectrum corresponding to sarcoplasmic reticulum was decomposed and the parameters obtained (position, width and gaussian fraction) used to produce the artificial curve. This artificial curve together with its deconvolved spectra using different apodization functions and K values, were submitted to the fitting procedure. The values corresponding to the parameters of the different components in both original and deconvolved curves are represented in Figure 1. The results corresponding to % band area are shown in Fig. 1A whereas Fig. 1B corresponds to the component bandwidths. Structure using Structure using only D,O spectra D,O and H20 spectra