Date Published: March 01, 2016
Publisher: International Union of Crystallography
Author(s): Vincent Olieric, Tobias Weinert, Aaron D. Finke, Carolin Anders, Dianfan Li, Natacha Olieric, Camelia N. Borca, Michel O. Steinmetz, Martin Caffrey, Martin Jinek, Meitian Wang.
The successful structure solution of the integral membrane diacylglycerol kinase and the CRISPR-associated endonuclease RNA–DNA complex by native SAD phasing is demonstrated. The structures were solved with a combined low-dose multi-orientation, multi-crystal data-collection strategy.
Native SAD (single-wavelength anomalous diffraction) is a de novo macromolecular structure-determination method in which the phase problem is solved by exploiting the anomalous diffraction signal of light atoms present in the crystal (for a review of anomalous diffraction, see Hendrickson, 2014 ▸). Unlike other de novo methods, structure solution by native SAD does not require the incorporation of exogenous heavy atoms, which can be tedious and can lead to non-isomorphism. However, native SAD phasing has its own challenges. Firstly, the absorption edge, where anomalous scattering is maximized, is below 2.5 keV (>5 Å) for elements such as sulfur and phosphorus, which is beyond the reach of most current macromolecular crystallography (MX) synchrotron beamlines. Data collection is also hindered by sample and air absorption of X-rays at such low energies. Thus, native SAD data collection is usually performed at ‘compromise’ energies of around 6 keV (2.066 Å; Mueller-Dieckmann et al., 2005 ▸; Liu et al., 2014 ▸; Weinert et al., 2015 ▸) to maximize the anomalous signal while minimizing absorption. Alternatively, home sources, which typically operate with Cu Kα (1.54 Å) or Cr Kα (2.29 Å) radiation, can also be used for such measurements. Secondly, because data collection is performed far from the absorption edges of these light elements, the resulting anomalous signal is small and accurate measurements of the reflection intensities and their differences are essential. To obtain this level of accuracy, random and systematic errors must be minimized. However, typical data-collection protocols at third-generation synchrotrons, usually performed around a single axis, yield anomalous data that can be adversely affected by radiation damage if insufficient care is given to the X-ray dose delivered to the crystal. Consequently, successful cases of phasing by native SAD remain rare; currently, about 150 native SAD structures have been deposited in the Protein Data Bank (Berman et al., 2000 ▸).
Native SAD could be considered the ‘ideal’ de novo phasing method for macromolecules because it relies solely on information from atoms that are naturally present in the sample. The method is readily applicable to challenging cases. Accordingly, there is every reason to try native SAD first with macromolecules that contain a sufficient number of naturally occurring anomalous scatterers. Here, we show that multi-crystal averaging of data sets used in conjunction with a low-dose, multi-orientation strategy is beneficial in difficult cases where data from one crystal are insufficient. Further, low-dose data collection in multiple orientations means that fewer crystal specimens are needed to provide the required data accuracy.