The use of anomalous scattering effects to phase diffraction patterns from macromolecules

Recent experiments using synchrotron radiation to measure anomalous scattering terms at wavelengths very close to L-shell elemental absorption edges [Phillips, Templeton, Templeton, & Hodgson (1978). Science, 201, 25 7-259] have demonstrated that effects much larger than those expected on the basis of simple atomic scattering calculations are observed. These observations of the large wavelength dependence of the magnitudes of anomalous scattering terms has prompted our reexamination of how such effects can be used to phase single-crystal diffraction patterns. This paper describes a methodology for using information on the magnitude of anomalous scattering effects to plan a multiple-wavelength phasing experiment on a macromolecule-containing crystal. Different data collection strategies, such as measuring data at many wavelengths less accurately or at a few wavelengths accurately, are compared. The method uses principles taken from standard MIR phasing techniques and also can be used to process multiple-wavelength data to obtain the phases. A numerical example of phasing, using the anomalous scattering curves determined experimentally for cesium, is presented and the results discussed in terms of the applicability of multiple-wavelength phasing for protein crystallographic studies. A survey of a variety of heavy metals suggests that large changes in anomalous scattering near the L absorption edges are quite general and will be of importance in multiple-wavelength phasing of single-crystal protein diffraction patterns.