ブックタイトル日本結晶学会誌Vol60No5-6

概要

日本結晶学会誌Vol60No5-6

Armin WAGNER，Ramona DUMAN，Kamel el OMARI，Vinay GRAMA，Vitaliy MYKHAYLYKHTable 1 Wavelengths and energies for elements of biological relevance in the tender X-ray range and their anomalouscontribution f”to the scattering factor for selected wavelengths.PSClKCaλ［A］5.77885.01554.39293.43693.0704E［keV］2.14552.47202.82243.60744.0381f”（1A）0.170.240.310.480.51f”（2A）0.710.901.121.682.01f”（3A）1.451.822.243.263.89f”（4A）2.342.903.550.550.68f”（5A）3.414.090.500.811.01reduced in the case of L or M absorption edges）. Thistechnique has recently been reviewed in Handing et al. 2）At low resolution, when side chains are not clearlyresolved, it can be very difficult to build the amino acidsequence into the electron density. Anomalous differenceFourier maps from a long-wavelength dataset can facilitatemodel building by revealing the position of the sulphursatoms from methionine and cysteine residues, providinganchor points to fit the sequence into the density.Experimental phasing by MAD（multi-wavelengthanomalous diffraction）and SAD（single-wavelengthanomalous diffraction）methods is a very powerfultechnique to overcome the crystallographic phase problemfor structure determinations of novel structures, particularlywhen no homology model for molecular replacement exists.Anomalous scatterers can be introduced by soaking, cocrystallisationor by replacing the amino acid methionineby seleno-methionine. 3）The selenium K absorption edgeis atλ＝0.98 A making this technique easily applicableat synchrotron beamlines, for those cases where proteinproduction and crystallisation are not adversely affectedby seleno-methionine incorporation. Ideally, the intrinsicanomalous signal from sulphur should be sufficient for nativephasing, however, with the sulphur K absorption edge atλ＝5.02 A, this signal is weak within the wavelength range ofstandard beamlines.While a first native phasing experiment from sulphur（S-SAD）was already described in 1981, 4）it was only in2000 when a first novel structure was published. 5）However,it took another 10 years until more challenging crystalstructures were reported 6-8）using standard MX beamlinesetups at wavelengths between 1.8 and 2.0 A.The absorption edges of the elements present in thenaturally occurring amino acids are outside the typicalwavelength range of synchrotron beamlines（Table 1）.While C, N and O are in the soft X-ray range, the edgesfrom phosphorus are in the so-called“tender X-ray regime”,still accessible with silicon based technology for bothmonochromators and detectors. However, the complexity isincreasing significantly when performing experiments withtender X-rays, or with wavelengths longer than 2 A. Thisis due to two physical effects. Firstly, the X-ray absorptioncross section increases approximately with the cube of thewavelength. Therefore, standard sample environments in airare no longer adequate, windows in the beam path have to beavoided and, at wavelengths close to the sulphur K absorptionedge, even the sample will absorb most of the incomingphotons, causing radiation damage. In addition, absorptioncorrections need to be seriously taken into account, to correctfor the different absorption lengths of individual reflections,as function of sample size and morphology. Figure 1a showsthe transmission through different thicknesses for lysozymecrystals as a function of wavelength, highlighting the effect ofabsorption.The second aspect to consider at long wavelengths isthe increase of diffraction angle. Bragg’s law describesthe relation between wavelength and diffraction angle 2θfor a given reflection（constant d-spacing）asλ～sinθ.Hence, diffraction angles increase significantly to longerwavelengths and standard flat detectors are no longer the mosteffective way to collect complete diffraction data（Fig.1b）.H. Stuhrmann performed first pioneering experiments inthe 1990s at wavelengths around the sulphur and phosphorusabsorption edges, however, neither the sample environmentnor the detectors available at this time allowed to perform asuccessful structure determination. 9）Synchrotron technology,in particular detectors, had to develop further until conceptsfor dedicated instruments optimised for long-wavelengthMX, providing easy access to wavelengthsλ＞3 A could bedeveloped and implemented. Currently, two such beamlinesexist, BL1A at Photon Factory, Tsukuba, Japan 10）and I23 atDiamond Light Source, Didcot, United Kingdom. 11）The long-wavelength MX beamline I23 is one of sevenbeamlines for macromolecular crystallography at DiamondLight Source. Beamlines I03, I04 and I04-1 mainly aimfor highest throughput applications, with I03 also offeringaccess to containment level 3 experiments. Beamline I24was the first dedicated tunable microfocus beamline for MX,234日本結晶学会誌第60巻第5･6号（2018）