RECENT DEVELOPMENTS IN MACROMOLECULAR TIME-RESOLVED CRYSTALLOGRAPHY

D.Bourgeois

LCCP, IBS, UPR 9015, 38027 Grenoble Cedex, France and ESRF, BP 220, 38043 Grenoble Cedex, France
email : bourgeoi@esrf.fr

Keywords : Protein crystallography, Structural intermediates, Cryotrapping, Laue technique.

Conformational changes form the basis of protein activity. The knowledge of rates constants and structures of reaction intermediates is therefore a key issue in fundamental mechanistic studies, and has practical applications such as in the field of drug design. Under physiological conditions, the lifetimes of structural intermediates span an extremely wide range of timescales, from fs to seconds or more. Whatever the timescale, visualization of a transient state by X-ray crystallography always relies on the ability to force a synchronous and homogeneous build-up of the population of interest within the crystal, and to maintain the latter until complete diffraction data have been collected. These two goals can be achieved with chemical (e.g. solvent trapping, use of photo-labile compounds), and/or physical (e.g. lower the temperature) methods, used in combination with standard monochromatic or fast (e.g. Laue), possibly stroboscopic, data collection techniques. In all cases, the authenticity of the observed intermediates has to be carefully assessed, and cross-validation using several techniques might be necessary. In the last few years, time-resolved protein crystallography has benefited from the development of both third generation synchrotron sources and flash-freezing techniques, allowing very fast timescales to be approached. However, several difficulties make this field challenging : the intermediates with shorter lifetimes are those whose extend of spatial modifications are fainter ; intermediate conformations usually have partial occupancy in the crystal ; they may show non canonical structures ; real time Laue experiments may be accurate but non precise due to severe signal to noise limitations ; low temperature structures may be precise but artefactual and therefore inaccurate. Definite conclusions about transient structures should take advantage of the complementarity between the two techniques and incorporate results from other spectroscopic or computational methods.

Based on recent experiments carried out at ESRF on s to ns timescales, these issues will be discussed.