Date Published: January 01, 2016
Publisher: International Union of Crystallography
Author(s): Chia-Ying Huang, Vincent Olieric, Pikyee Ma, Nicole Howe, Lutz Vogeley, Xiangyu Liu, Rangana Warshamanage, Tobias Weinert, Ezequiel Panepucci, Brian Kobilka, Kay Diederichs, Meitian Wang, Martin Caffrey.
A method for performing high-throughput in situ serial X-ray crystallography with soluble and membrane proteins in the lipid cubic phase at cryogenic temperatures (100 K) is described. It works with nanogram to single-digit microgram quantities of protein and lipid (and ligand when present), and is compatible with both high-resolution native data collection and experimental phasing without the need for crystal harvesting.
The lipid cubic phase (LCP) or in meso method for crystallizing membrane proteins has delivered over 290 structures of integral membrane proteins (Caffrey, 2015 ▸). This represents about 12% of the membrane-protein structures in the Protein Data Bank (PDB). Importantly, half of these PDB records have been deposited in the last two years, indicating that the method is experiencing explosive growth. It is responsible for some of the highest profile structures in the field, particularly in the area of G-protein-coupled receptors (GPCRs). Further, the method works with a wide range of membrane-protein types, sizes and sources, including transporters, channels, peptides and enzymes. Soluble proteins can also be crystallized in the lipid cubic phase. A recent application for the mesophase is as a medium in which to perform injector-based serial crystallography with free-electron lasers (FELs; Liu et al., 2013 ▸; Caffrey et al., 2014 ▸; Weierstall et al., 2014 ▸; Fenalti et al., 2015 ▸; Kang et al., 2015 ▸; Li et al., 2015 ▸) and synchrotron X-rays (Botha et al., 2015 ▸; Nogly et al., 2015 ▸).
The success of the IMISXcryo approach to high-throughput structure determination for crystals grown by the in meso method without the need for the time-consuming and inefficient crystal-harvesting step is illustrated in this section. We begin by demonstrating that the method works with the test soluble proteins insulin and lysozyme. In the case of insulin, structures were solved by MR and by sulfur SAD phasing. For lysozyme, both MR and bromine SAD were used. The integral membrane proteins in the study included AlgE, PepTSt, β2AR and DgkA, and MR phasing was used with all four. The point of this paper is not to describe the structures but rather to demonstrate that the IMISXcryo method works easily and efficiently with a range of protein types and that it can be used both for high-resolution native data collection and for experimental phasing.