The rise of neutron cryo-crystallography

Abstract

The use of boiled-off liquid nitrogen to maintain protein crystals at 100 K during X-ray data collection has become almost universal. Applying this to neutron protein crystallography offers the opportunity to significantly broaden the scope of biochemical problems that can be addressed, although care must be taken in assuming that direct extrapolation to room temperature is always valid. Here, the history to date of neutron protein cryo-crystallography and the particular problems and solutions associated with the mounting and cryocooling of the larger crystals needed for neutron crystallography are reviewed. Finally, the outlook for further cryogenic neutron studies using existing and future neutron instrumentation is discussed.

Keywords: cryogenic data collection; enzyme mechanisms; neutron crystallography.

Figure 1

The neutron structure of DHFR in complex with folate and NADP⁺. (a) The neutron structure (Wan, Kovalevsky et al., 2014; Wan, Bennett et al., 2014 ▸). N atoms are shown in blue, C atoms in yellow, S atoms in gold and O atoms in red. Nuclear density 2F _o − F _c maps are represented at a σ level of 0.8. (b) The X-ray structure collected at 100 K (Wan, Kovalevsky et al., 2014 ▸). C atoms are shown in cyan and water molecules are represented by red spheres. Electron density 2F _o − F _c maps are represented at a σ level of 1.2. (c) Structure alignment of the neutron structure (Wan, Bennett et al., 2014 ▸), the 100 K X-ray structure (Wan, Kovalevsky et al., 2014 ▸), the 277 K X-ray structure (Wan, Bennett et al., 2014; C atoms represented in orange) and the neutron structure of DHFR in complex with methotrexate (Bennett et al., 2006; C atoms represented in green). For clarity, density is shown only for certain active-site residues and ligands.

Figure 2

(a) A cryocooled perdeuterated crystal of Toho1 β-lactamase in complex with the antibiotic cefotaxime, freeze-trapping a short-lived intermediate over a 40 h data-collection time. The crystal volume is around 0.5 mm³ and it is mounted in a specially made 2 mm diameter litholoop from Molecular Dimensions. The large crystal fits snugly into the custom-sized loop, making it less likely that the crystal will be lost during the cryocooling process. (b) A time-of-flight wavelength-resolved slice from one of five 8 h diffraction images from a Toho1 β-lactamase data collection on the MaNDi instrument.

Figure 3

The structures of the intermediates of haem peroxidases. Nuclear density is shown in green and difference density for the ligand is shown in black. (a) The ferric (resting) enzyme cytochrome c peroxidase, (b) compound I of cytochrome c peroxidase, (c) compound II of ascorbate peroxidase.

Figure 4

The intermediates in the reaction of haem peroxidases. The structures determined by NPX are circled. (a) The ferric (resting) enzyme from cytochrome c peroxidase, (b) compound I of cytochrome c peroxidase, (c) compound II of ascorbate peroxidase. The radicals on the porphyrin in the case of ascorbate peroxidase and on a tryptophan residue in the case of cytochrome c peroxidase are shown as and , respectively. The red crosses show alternative structures eliminated by NPX.