Show Abstract ▼

Abstract: Since its first use to solve the structure of sodium chloride in 1915 X-ray crystallography has developed significantly to become the premier technique for obtaining 3D structural information of small molecules and macromolecules alike. As the technique continues to develop and focus its attention on weak diffraction from the likes of micro-crystals and poorly packed crystals of membrane proteins and large protein complexes; as well as ultra-high resolution data and weak anomalous signal from native atoms, data quality is becoming more and more important. Data quality is particularly important in the wake of long wavelength macromolecular crystallography (MX) for phasing using anomalous signal from native sulphur and phosphorous atoms in proteins and DNA. This thesis first investigated the use of a new sample handling technique using a humidity controlled stream to preserve macromolecular crystals while excess surrounding solvent is removed (Chapter 2). Following the successful development of this technique the effects of excess surrounding solvent on data quality was assessed when collecting at standard MX X-ray wavelengths (~ 1 Å) and longer X-ray wavelengths (~ 2 Å). Datasets were collected from large populations of control and test crystals at standard and longer wavelengths to allow robust statistical methods to be applied; a practice not widely adopted in method development studies in X-ray crystallography. This made it possible to assess the small differences in data quality in the presence and absence of excess surrounding solvent. The effects of surrounding solvent at longer wavelengths appear to be protein dependent with some proteins tested showing no significant difference and others a significant decrease in data quality at longer wavelengths (Chapter 3). Originally this project aimed to use the new long wavelength in-vacuum MX beamline, I23, at Diamond Light Source UK to carry out phasing experiments using native sulphurs for structure solution. However, the considerable complexity involved in developing in-vacuum MX meant these experiments could not be carried out during the time frame of this thesis. Chapters 4 and 5 outline the production of a novel cancer protein (cancerous inhibitor of protein phosphatase 2A) and two protein targets from the Achromobacter xylosoxidans (Ax) genome intended for sulphur single wavelength anomalous dispersion phasing experiments on I23. Of these proteins the structure of Ax-α/β hydrolase was solved by conventional methods, the structure of which is discussed in Chapter 5. Of the protein crystals used in long wavelength data quality experiments in Chapter 3 the molecular biology of PA3825-EAL, a biofilm regulating protein essential to the swarming ability of Pseudomonas aeruginosa, was investigated further. The crystal structure of PA3825-EAL was solved in the resting, substrate bound and product bound states to high resolution. Comparison of the crystal structures of monomeric and dimeric PA3825-EAL with the inactive dimeric structure of MucR-EAL suggests dimerisation via helix 8 plays a role in inhibition of EAL domains. Prior to this, dimerisation was thought to be an activating factor in EAL domains. The product bound state of PA3825-EAL showed the presence of a previously unreported third metal binding site which may form an essential component of the reaction mechanism of EAL domains. Inability of MucR-EAL to incorporate this third metal due to dimerisation may explain the lack of activity despite possessing the conserved catalytic residues necessary. The fast detector technology and improvements in automated data processing software that allowed diffraction data for large populations of crystals to be collected in Chapters 2 and 3 have also been applied to development of a serial data collection technique.