Lipidic cubic phase serial millisecond crystallography using synchrotron radiation.

Przemyslaw Nogly,Cecilia Wickstrand,Richard Neutze,Daniel James,Manfred Burghammer,Markus Metz,Petra Båth,Wenting Wu,Anastasya Shilova,Garrett Nelson,Vadim Cherezov,Peter Berntsen,Dingjie Wang,Isabel Moraes,Joerg Standfuss,Thomas A White,Haiguang Liu,Kathrin Jaeger,Uwe Weierstall,Michaël Heymann ,D Oberthüer ,Nadia A Zatsepin ,Henry N Chapman ,Valérie Panneels ,Gebhard F X Schertler ,Linda C Johansson ,J David Spence

doi:10.1107/s2052252514026487

Abstract

Lipidic cubic phases (LCPs) have emerged as successful matrixes for the crystallization of membrane proteins. Moreover, the viscous LCP also provides a highly effective delivery medium for serial femtosecond crystallography (SFX) at X-ray free-electron lasers (XFELs). Here, the adaptation of this technology to perform serial millisecond crystallography (SMX) at more widely available synchrotron microfocus beamlines is described. Compared with conventional microcrystallography, LCP-SMX eliminates the need for difficult handling of individual crystals and allows for data collection at room temperature. The technology is demonstrated by solving a structure of the light-driven proton-pump bacteriorhodopsin (bR) at a resolution of 2.4 Å. The room-temperature structure of bR is very similar to previous cryogenic structures but shows small yet distinct differences in the retinal ligand and proton-transfer pathway.

Highlights

Structure determination by X-ray crystallography has developed continuously over the last century, yielding structures of ever more difficult and complex molecules
The key advantages of this method are: (i) crystal injection using the Lipidic cubic phases (LCPs) combined with a microfocus beamline allows diffraction data to be collected at room temperature, and crystal freezing and difficult crystal handling steps such as mounting crystals in a loop are not necessary; (ii) thousands of crystals can be screened in a short time and with less than a milligram of protein; (iii) microfocus beams at storage-ring sources are widely available and beam access is unlikely to limit SMX; and (iv) the method is well suited for time-resolved diffraction studies on the microsecond to millisecond timescale
The synchrotron cryocooled bR structure was solved using this study, we have adapted the technology for use at more molecular replacement with ground-state bR

Summary

Introduction

Structure determination by X-ray crystallography has developed continuously over the last century, yielding structures of ever more difficult and complex molecules. A gas dynamic virtual nozzle (GDVN) (DePonte et al, 2008; Weierstall et al, 2012), which was the injection device for these first experiments (Chapman et al, 2011; Boutet et al, 2012; Johansson et al, 2012), can deliver crystals in their lowviscosity crystallization buffer/mother liquor at a liquid flow rate of about 10 ml minÀ1 and a speed of about 10 m sÀ1 At this flow rate, and with the repetition rate of the hard XFEL sources currently in operation, most crystals flow past the interaction point in the time between X-ray pulses and are wasted. At room temperature, and crystal freezing and difficult crystal handling steps such as mounting crystals in a loop are not necessary; (ii) thousands of crystals can be screened in a short time and with less than a milligram of protein; (iii) microfocus beams at storage-ring sources are widely available and beam access is unlikely to limit SMX; and (iv) the method is well suited for time-resolved diffraction studies on the microsecond to millisecond timescale

Purification

Crystallization

Model building and refinement

Sample preparation and LCP injection

Findings

Discussion