Date Published: March 01, 2019
Publisher: International Union of Crystallography
Author(s): Mostafa Jamshidiha, Inmaculada Pérez-Dorado, James W. Murray, Edward W. Tate, Ernesto Cota, Randy J. Read.
The solution of a structure of human Rab27a suffering from severe anisotropy and translational noncrystallographic symmetry was aided by identifying diffraction measurements with low information content.
Accounting rigorously for the effects of errors in a statistical model can dramatically enhance the sensitivity of likelihood-based methods. For instance, in molecular-replacement (MR) calculations, Phaser (McCoy et al., 2007 ▸) is able to account for the effects of errors in both the search model and in the measured diffraction data; this is difficult to achieve with methods based on the properties of the Patterson function or on the computation of correlation coefficients. In addition, information obtained from already placed search components significantly improves the signal in rotation and translation searches for subsequent components, as measured by the log-likelihood gain (LLG) and Z-scores (McCoy, 2007 ▸; Storoni et al., 2004 ▸; McCoy et al., 2005 ▸).
The hRab27aMut(GppNHp) data show how difficult cases of molecular replacement can be solved using Phaser if anisotropy and tNCS are properly accounted for using strategies that are applied automatically in Phaser v.2.7.17 or newer. Moreover, the structure of the hRab27aMut(GppNHp) crystals shows that the SF4 pocket, which is the primary target for ligand-binding studies, is unoccupied and could be used to study the structure of ligands binding to Rab27a. The only major drawback is the data quality, specifically the overall resolution and severe anisotropy, which would be problematic for weak binding ligands with low occupancy. Optimization of crystallization conditions, additive screens and the structure of hRab27aMut(GppNHp) reported here will guide further construct design to obtain a more tractable crystal form for ligand-binding studies.