DATA COLLECTION WITH R-AXIS II, AN X-RAY DETECTING SYSTEM USING IMAGING PLATE

Transcription

1 THE RIGAKU JOURNAL VOL. 7 / NO. 2 / 1990 Technical Note DIFFRACTION DATA COLLECTION WITH R-AXIS II, AN X-RAY DETECTING SYSTEM USING IMAGING PLATE ATSUSHI SHIBATA R&D Division, Rigaku Corporation 1. Introduction It is well known that a two-dimensional (2D) detector is efficient in collecting a number of X- ray diffraction data from macromolecular crystals, e.g. protein crystals. Various detecting methods have been proposed, some of which have been practically used. Among those detectors, an imaging plate [1] (hereafter referred to as IP) has superior fundamental characteristics as an X-ray detector in respect of sensitivity, linearity, dynamic range, etc, and therefore, is going to be in active use as a 2D detector collecting diffraction data. Although, 2D detector system using an IP has been already outlined in a previous issue of this journal [2] significant improvement have been made since then. Here we describe a new model R-AXIS II briefly, along with some results of diffraction measurements from protein crystals. 2. System Outline The R-AXIS (Rigaku Automated X-ray Imaging System) model II, shown in Fig. 1, is a fully automated rotation camera system for macromolecular crystallography, which uses the IP as a 2D detector in place of conventional X-ray films. Fig. 2 shows the X-ray optics and the detecting section equipped with the IP's of the system. Since the characteristics of the IP have been already published [3, 4], we concentrate only on the detecting mechanism and the scanning method a doped in the system. Incidence of X-rays produces color centers in the IP where the X-ray photons hit, making a latent image on the IP. When the IP having such color centers is irradiated by a He-Ne laser beam (wavelength 633 nm), the beam is absorbed by the color centers, stimulating luminescence with a peak at 390 nm and restoring them to the ground state. By scanning the laser beam and detecting the luminescence simultaneously over the IP plane, the two-dimensional latent image kept in the IP is detectable. The residual color centers in the IP after reading procedure Fig. 1 R-AXIS II system combined with a rotating anode X-ray generator. are erased completely by irradiation with visible light. This makes it possible to use the IP repeatedly, and hence to make an automated measurement system. As shown in Fig. 2, two types of X-ray optics are available for the R-AXIS II system: a) a flat crystal monochromator (usually graphite) with a parallel beam collimator, or b) point focusing optics using double focusing mirrors. In the former optics, the spatial resolution of diffraction spots is inferior to the latter owing to no focusing mechanism, but the optics is relatively compact in size and optical alignment is easier. The latter optics has a higher spatial resolution and is suitable for measurement of crystals having large unit cells. Nonetheless, experiments have shown that crystals with cell length up to ca. 150 Å are measurable even by using the former optics. In this system, a rotary shutter is designed to respond at high speed, synchronized with the φ 28

2 Fig. 2 Configuration of X-ray optics and IP-reading unit of R-AXIS II. rotation. This makes it possible to collect integrated intensities of partially recorded reflections without lowering accuracy. By synchronizing the rotary shutter with an attenuator and a beam stopper, the direct beam is recorded (origin recording) whenever the frame exposure is completed. A pair of IP s is used in the system so that, during exposure of one IP, the other IP at the opposite side can be read simultaneously. This alternative exposure system is especially useful when decay of a crystal is considerable. The IP moves from the exposure position to the reading position by means of a 180 rotation of a base supporting the two IP's. At the midpoint of the rotation, the movement stops for ca. 20 sec to expose the other IP to a light, in order to erase the residual color centers in the IP after reading. An objective lens for readout scans horizontally over the IP surface at a constant speed by means of a linear motor. Synchronized with the horizontal scanning, a stage, on which the linear motor and a detector (photomultiplier tube) are mounted, moves vertically. A highly accurate two-dimensional scanning in an orthonormal coordinate system is achieved by a combination of these mutually perpendicular motions. The size of the IP is 200 x 200 mm 2 and the size of a pixel is 105 x 105 µm 2. The frame size of 1900 x 1900 pixels is approximately 8 M bytes (2 bytes/pixel). Fig. 3 shows the correlation between the number of incident X-ray photons (monochromatized CuKα radiation) and the output (the value after analogue-to-digital conversion) of the R- AXIS II system. It is clear that the system has excellent linearity and a wide dynamic range. Fig. 4 shows an example of a diffraction pattern obtained from a hen egg white lysozyme crystal. Graphic software displays a gray scale representation of a frame on the CRT. The figure can be magnified at any rate and hardcopy is available on a laser printer, although the quality of the hardcopy image is less than that of the figure shown on the CRT. The diffraction pattern shown in Fig. 4 is one of such hardcopies. Fig. 3 Linearity and dynamic range of R-AXIS II response for CuKα radiation. Vol. 7. No

3 Initial setup with data recall which has been keyed-in at R-AXIS operation stage Determination and refinement of crystal orientation, based on still frames Confirmation of crystal orientation, etc. by plotting frame on a graphic display. Evaluation of integrated intensities from oscillation frames Fig. 5 shows a block diagram of the standard R-AXIS II system. A VAX 3100 workstation was selected as the host computer that controls the system, taking into account that VAX computers are used worldwide by crystallographers and various other useful software is readily available. Data processing is also carried out on the workstation. 3. Data processing Data processing software extracts reflection data from a two-dimensional frame obtained by the R-AXIS II system. A flow chart of the processing procedure is given in Fig. 6. By adopting a 30 Fig 4. An example of a diffraction pattern of protein crystals measured on R-AXIS II. Sample: hen egg lysozyme crystal, tetragonal, space group P , a=79.1, c=37.9 Å; X-ray beam approximately parallel to c-axis; Crystal-to-IP distance 90 mm; IP size 200x200mm 2 ; CuKα radiation monochromatized by graphite monochromator with parallel beam collimator; Crystal rotation range 5 ; Exposure time 30 min. Fig. 5 Block diagram of R-AXIS II system. Calculation of scale factors of frames, averaging intensities of symmetryrelated reflections, analysis of anomalous dispersion effect, and final output of independent reflection data on ASCII file. Fig. 6 Flow of data processing procedure. menu-driven operation system throughout, a user can operate the software without any difficulty. The main features of the processing software are as follows: (1) Auto-indexing procedure [5], by which crystal orientation is automatically determined from still frames (2) Rossmann's profile fitting method to evaluate integrated intensities (3) Parallel data processing, that data processing and data acquisition can be carried out almost simultaneously (4) Various graphic functions (5) Analysis of an anomalous dispersion effect (6) Final output of independent and scaled reflection data (h, k, l, F o ) 4. Measurement of Protein Crystals Table 2 shows a summary of measurement of some protein crystals, using CuKα radiation from a 0.3 mm 2 fine focus rotating anode (40 kv, 100 ma) with an optics of a graphite monochromator and a parallel beam collimator. It must be noted that, in each measurement, a set of reflection data was collected by using only one crystal. The R merge 's, a statistical measure of internal consistency, were 4.20, 4.19, 4.87, and 5.27%. These values are satisfactorily small, indicating that good-quality diffraction data from protein crystals can be obtained on the R-AXIS II system. Fig. 7 shows the distribution of Rmerge 's of a hen egg white lysozyme crystal in various resolution ranges. In case of partially recorded reflections, a full-integrated intensity is calculated by

4 Readout unit Planar two-dimensional scanning system X-ray exposure/readout parallel run Table 1 Major Specifications of R-AXIS II Excitation beam : He-Ne laser Fluorescence detector : Photomultiplier tube Pixel size : 105 x 105 µm 2 Pixel number : 1900 x 1900 Detection area (IP size) : 200 x 200 mm IP : 2 plates Output data : 7.2 MB frame Beam stopper : Fixed to 6µ mylar, automated retreat at time of direct beam photography X-ray optical system Flat crystal monochromator Monochromator : Graphite optics Collimator : 0.3φ double collimator 0.5φ single collimator Attenuator : Automatic setting at time of direct beam photography φ-axis goniometer Minimum feed : step (stepping motor driving) Manual rotation : 360 free rotation Goniometer head : IUCr standard Eucentric type Telescope : 40-time magnification Bench for readout unit Crystal-to-IP distance : 50 ~ 200 mm (with 0.1 mm vernier) Max. detection angle (2θ max) : 61.2 (d min =1.51Å) /CuKα Controller/Driver With funtion of X-ray exposure, IP data readout : High-speed 16 bits ADC IP data readout and Interface for control : RS-232C data transfer Data transfer interface : SCSI Host computer VAX 3100 CPU : KA42-AA 90 nsec with 32 KB cache Memory : 16 MB Fixed disk : 323 MB Backup : TK50 95 MB CRT : 19 color monitor Laser printer : 300 dpi RS232C/SCSI interface provided Software Crystal axis alignment/still measurement/occ Measurement/Data processing (Orientation Refine index/f output) Graphic : Graphic display of IP data file on CRT. Support for axis alignment, etc. can be made with different cursors. adding two adjacent partial intensities recorded on the two contiguous frames. R merge 's of fully recorded reflections and those of partially recorded reflections are plotted separately in Fig. 7, to assess the accuracy of thus evaluated values of partially recorded intensities. Their agreement is good, especially in lower resolution ranges. This indicates, even though indirectly, that the φ-axis rotation mechanism, including the highly responsive rotary shutter incorporated in the optics, does function successfully. In general, the anomalous dispersion effect is so weak that it is not expected to be observed without using a high-precision X-ray detector. In order to examine whether the R-AXIS II system is able to measure the anomalous dispersion effect, the following experiment was carried out. A sample crystal of hexagonal myoglobin (space group P 6, a = 91.20, c = Å) was provided by Prof. G. Phillips of Rice University. An iron atom in myoglobin, of which the K absorption edge is 1.73 Å, is an anomalous scatterer for CuKα radiation (1.5418Å) with the anomalous dispersion term f" = 3.4e. The experimental condition was: X-ray source: Cu Kα radiation monochromatized by graphite monochromator Crystal orientation: c-axis approximately parallel to the spindle axis Oscillation range: 1 /frame Total frame number: 60 Vol. 7. No

5 Table 2 Typical RAXIS-II measurement Crystal Lysozyme IPMDH1 IPMDH2 Crystal-A Crystal System Tetragonal Trigonal Orthorhombic Space Group P P3221 a (Å) b (Å) c (Å) Resolution (Å) Crystal-to-IP distance (mm) Exposure time (min/frame) Oscillation range per frame ( ) Total oscillation range ( ) Number of frames R merge (%)* R merge (%)* of full reflections R merge (%)* of partial refls Number of total observations Number of rejected observations No. of independent reflections Yield (%) All the measurements were carried out using CuKα radiation (40kV, 100 ma) with graphite monochromator * Rmerge(%) = 100 I Gi I I, where G i is an inverse scale factor. Fig. 7 Distribution of R merge s After routine data processing, independent reflections up to 1.7 Å resolution were obtained (completeness: 78.9%), with an R merge of 6.00%. The anomalous dispersion effect was analyzed using reflections with F 2 >2, and 4672 significant Bijvoet pairs [ anom =F(+)-F(-)] were obtained. On the Harker section of the anomalous difference Paterson map, of which coefficients are 2 anom, Fe-Fe self vectors appeared as the highest peaks, as shown in Fig. 8. Fig. 8 Harker section (w=0) of myoglobin anomalous difference Patterson map. Coutours are drawn at 1, 2 and 3σ. References and Notes [1] Imaging Plate: Produced by FUJI PHOTO FILM CO. LTD. [2] "X-Ray Diffractometer System For Macromolecules, R- AXIS II.", Vol. 5, No.1, (1988), 34. [3] N. Kamiya, Y. Amemiya andj. Miyahara:J. Crystallgr. Soc. Japan, 28 (1986), 350. [4] Y. Amemiya, Y. Sato, T. Matsushita, J. Chikawa, K. Wakabayashi and J. Miyahara: Topics in Current Chemistry, 147 (1988), 121. [5] T. Higashi: J. Appl. Crystallogr., 23 (1990),