Pink-beam serial crystallography

Abstract

Serial X-ray crystallography allows macromolecular structure determination at both X-ray free electron lasers (XFELs) and, more recently, synchrotron sources. The time resolution for serial synchrotron crystallography experiments has been limited to millisecond timescales with monochromatic beams. The polychromatic, "pink", beam provides a more than two orders of magnitude increased photon flux and hence allows accessing much shorter timescales in diffraction experiments at synchrotron sources. Here we report the structure determination of two different protein samples by merging pink-beam diffraction patterns from many crystals, each collected with a single 100 ps X-ray pulse exposure per crystal using a setup optimized for very low scattering background. In contrast to experiments with monochromatic radiation, data from only 50 crystals were required to obtain complete datasets. The high quality of the diffraction data highlights the potential of this method for studying irreversible reactions at sub-microsecond timescales using high-brightness X-ray facilities.

Fig. 1

Exemplary single shot pink-beam diffraction images of a proteinase K and b phycocyanin microcrystals. Data were recorded at the BioCARS beamline using a Rayonix CCD detector. Crystal sizes were between 10 and 20 µm for proteinase K and 30 and 40 µm in case of phycocyanin. The diffraction images reveal the absence of a pronounced water ring that is typically observed in diffraction experiments from macromolecular crystals. c, e Insets of subfigures a and b, respectively, highlighting the shape of the Bragg spots. d, f Lineouts of the red traces in c and e, respectively, showing the shape of two exemplary Bragg spots from proteinase K and phycocyanin diffraction patterns

Fig. 2

Quantitative analysis of the scattering background. Background scattering level from phycocyanin crystals mounted on a silicon chip (excluding diffraction intensities from Bragg peaks) highlighting the extremely low background level achievable with 3 × 10¹⁰photons/pulse (a). Radially averaged background level for the diffraction image as function of resolution (b)

Fig. 3

Comparison of electron density maps from proteinase K structures obtained from different diffraction methods showing the calcium binding site with the coordinating water molecules. a Structure obtained from our single shot pink-beam serial crystallography experiment. b Proteinase K structure from conventional single crystal rotation photographs (PDB ID: 2PRK). The blue grid represents 2mF_o− DF_c maps (left side) and 2mF_o− DF_c simulated annealing composite omit maps (right side) both at a contour level of 1.5σ, respectively

Fig. 4

Overall structure of the biological assembly of phycocyanin: symmetry-related chains A in cyan, chains B in magenta displayed as cartoon. The phycocyanobilin molecules are shown in green. a View along the symmetry axis with b being a 90° rotation displaying the assembly from the side

Fig. 5

Difference electron density maps of different phycocyanin datasets. To assess the influence of phase bias introduced by the search model, we mutated residues 152 and 153 of chain A of Phyco A1 (single chip/single pulse, a, pink), Phyco B (five chips merged, single pulse, e, cyan), and Phyco C (single chip, multi-bunch exposure, i, green) and calculated simulated annealing maps (all blue maps are 2mF_o− DF_c maps at 1.5 sigma, all green and red maps are mF_o− DF_c maps at 2.5 sigma). We furthermore deleted residues 147–157 of chain B for all three cases and calculated a simulated annealing omit map (b, f, j). Here we also show the omitted residues for clarity. In all three cases strong positive difference electron density is visible in the map in place of the omitted residues. To show that the mode of data collection does not lead to local radiation damage, we show the 2mF_o− DF_c map (c, g, k) and the mF_o− DF_c map (c, g, k) and (d, h, l) for the bond between Cys84 of chain A and the phycocyanobilin-ligand for all three cases. No difference electron density is visible around the S–C-bond in any of the cases

Fig. 6

Comparison of exemplary electron density maps of phycocyanin (2mF_o− DF_c at 1.5σ level) showing one of the phycocyanobilin molecules present in the structure (left) and a view along residues 90–95 of α-helix 6 (right) obtained by pink-beam serial crystallography at a synchrotron (a–c) and with serial femtosecond crystallography (d). a The electron density of dataset Phyco_A1 obtained from the measurement of 45 diffraction images using the APS single bunch. b The electron density of dataset Phyco_B obtained from the measurement of 205 diffraction images using the APS single bunch. c The same regions obtained from 52 diffraction images using a longer exposure time of 3.68 μs per crystal. d Structure 4ZIZ from the PDB based on 6679 single shot diffraction images collected at LCLS

Fig. 7

a Experimental setup installed at the BioCARS instrument at the APS for fixed target serial crystallography using the pink beam. The micro-patterned silicon chip is raster-scanned through the X-ray beam using the goniometer installed at the beamline. An inline microscope using Schwarzschild optics is used for sample visualization. It provides the opportunity of through the lense illumination for both sample visualization and laser excitation of the sample. b Close-up showing the chip mounted on the beamline goniometer and being exposed to a stream of humidified helium gas preventing the crystals from drying out. c Schematic view of phycocyanin crystals mounted onto a micro-patterned silicon chip. The short path of the direct beam between the collimator tube and the capillary beamstop behind the sample is highlighted in pink. By flushing the remaining free beam path with helium, extremely low background scattering levels are achieved. The drawings were created using the software Solid Edge ST8, KeyShot 5.0, GIMP 2, and Microsoft PowerPoint

Fig. 8

Evolution of collimator and beamstop designs. a Collimator and beamstop design typically used at protein crystallography beamlines at second generation synchrotron sources. A fraction of a relatively large X-ray beam is selected by using a 0.5 mm inner diameter collimator tube extending 5 mm close to the sample. After interaction with the sample, the direct beam is blocked with a 2 mm outer diameter metal cylinder with a blind hole placed about 30 mm behind the sample. The total path of the direct beam in air is about 35 mm. b Beamstop design currently used at most third generation synchrotron protein microfocus crystallography beamlines where much smaller samples can be investigated. Owing to the significantly smaller X-ray beam waist, the outer diameter of the beamstop can be one millimeter or less. Placed at a distance of again 30 mm low q-reflections can be now recorded up to a half opening angle of 16.7 mrad. The total path of the X-ray beam in air is again 35 mm. c Low background collimator and beamstop concept used for the present work. The path of the direct beam in air is reduced to 8 mm. X-ray photons scattered by air within the capillary are absorbed by the capillary walls. Compared to previous beamstop concepts a, b the path of the direct X-ray beam in air is shortened to 8 mm—reducing the background scattering level by a factor of about 4. At the end of the capillary the direct beam is blocked with a solid absorber. d Future beamstop concept: already 4 mm behind the sample the direct beam is fully enclosed by a 0.25 mm outer and 0.15 mm inner diameter metal capillary. The direct beam can be either blocked with a solid absorber or the tube can be extended through a hole in the detector and then blocked behind (favorable for XFEL applications). For data collection at cryogenic temperatures, collimator and post-sample beam-capillary would require heating, to prevent ice deposition caused by the cold-stream