Given a protein which contains a heavy atom possessing a significant white line at its absorption edge it has been shown that it is possible to perform an experiment at highly optimised X-ray energies which provide the possibility of solving the heavy atom structure and allow determination of structure factor phases. In particular it was shown that a protein structure of molecular weight 14600 could be solved by the preparation of an Iridium derivative and the measurement of data at three optimised X-ray energies. The structure contained six Ir sites with occupancies between 0.2 and 0.56.
Measurement of diffraction data at the rising and falling edges of the white line, where the values of were similar but a difference of 13.6e existed between their values of , permitted straightforward determination of the heavy atom positions using direct methods and also contributed to the phase solutions with the largest correlation coefficients compared with the target electron density. Measurement at these two energies and a data set at a third energy, either at the white line maximum or at a remote X-ray energy above the absorption edge, yielded the best combination of high correlation coefficient and minimum amount of data collected. The measurement of such data requires the use of a narrow band pass tunable X-ray source which has stability of its X-ray energy to better than . A bandpass of was used during these experiments but a slightly broader bandpass could be used without a significant change in the values of and . If data is to be measured at X-ray energies away from the absorption edge where anomalous scattering factors are approximately constant over energy ranges of one could contemplate making those measurements on a broader bandpass beam line where the higher intensities could imply shorter data collection times or even on a conventional source.
Heavy atoms which are suitable for performing such optimised experiments are the Lanthanides and the metals of the 3rd transition series from Hf up to and including Ir. Fluorescence measurements on the transition series metals showed a trend in the maximum height of the white line at the edge having a maximum at Hf and decreasing as the atomic number increases. The height and width of the white line are a function of the number of unoccupied states in the 5d band and the width of band. Because the transition series metals have their edges at higher X-ray energies than those of the Lanthanides they are preferable for MAD work due to the reduced effect of radiation damage at higher X-ray energies. It is also advantageous to select heavy atom compounds which have octahedral or tetrahedral coordination as this provides an isotropic environment resulting in isotropic anomalous scattering effects. In addition heavy atom compounds which are specific in their binding to proteins and which bind strongly with high occupancies are preferable. There are presently few choices available which are known to satisfy all of these conditions. Their is a need for a study of the chemistry of suitable heavy atom compounds in an attempt to isolate a set of the best candidates which suit particular protein solvent conditions.
The results produced by the phasing program MLPHARE were found to agree well with experimentally determined values of anomalous scattering factors and produced electron density maps which were considered to be of a good enough quality to allow interpretation of the protein main chain. The figures of merit produced by this software showed some correlation with the number of data sets used for the phase determination and therefore sometimes gave an over optimistic impression of the quality of the solution.