Deutsche Tagung f ¨ur Forschung mit ... - SNI-Portal

Biologische Systeme und Medizin Poster: Mi., 14:00–16:30 M-P210
The Use of Longer X-ray Wavelengths in Macromolecular Crystallography
Manfred S. Weiss 1 , Christoph Mueller-Dieckmann 2 , Santosh Panjikar 1 ,
Paul Tucker 1
1 EMBL Hamburg Outstation, c/o DESY, Notkestr. 85, D-22603 Hamburg, Germany
– 2 ESRF, 6 Rue Jules Horowitz, F-38043 Grenoble Cedex, France
The use of longer X-ray wavelengths (λ = 1.5-3.0 ˚A) in macromolecular crystallography
has over the past few years become a tool for phase determination using the anomalous
signal derived from the natively present sulfur and/or phosphorus atoms. In order
to extract the maximum possible information contained in the data, it is essential,
however, to apply a proper scaling protocol to the data. The best protocols in this
respect take absorption into account implicitly. Nevertheless, there seems to be an
optimal wavelength at which the anomalous signal-to-noise ratio is highest. In an
extensive examination of 75 diffraction data sets collected at different wavelengths
from ten different protein and DNA crystal systems, it was found that irrespective of
the anomalously scattering substructure present, this optimal wavelength was around
2.0-2.1 ˚A.
A further use of a well measured long wavelength data set is in the unequivocal
definition of the anomalously scattering substructure. Although a large percentage of
proteins are crystallized in the presence of ions derived e.g. from NaCl, KCl, Na2SO4
and Na3PO4 only about 10-15 % of all PDB entries contain such ions as part of the
structure coordinates. In order to further investigate this seeming contradiction, we
examined the substructures of 23 different protein crystals using a wavelength of λ =
2.0 ˚A. In all but two of the cases, we were able to identify additionally bound ions
for these proteins. Our findings thus suggest that a data set collected at a longer
wavelength appears to be essential in order to completely describe a protein structure.