 |
Just prior to the end of the last running period, the protein crystallography group just received a Quantum-4 CCD detector system from Area Detector Systems Corp. (ADSC). This detector employs 4 individual CCD detector chips each linked to an X-ray phosphor using a tapered fibre-optics coherent light guide. The four individual modules are arranged in a 2 x 2 matrix that has a total active area of 188 mm x 188 mm, with a total of 2304 x 2304 pixels that are each 82 µm x 82 µm. During an initial trial on beamline 1-5AD, our tunable-wavelength bending magnet beamline used for MAD experiments, it performed very well. This detector has the ability to read out an entire high resolution diffraction image in approximately 9 seconds, and will clearly dramatically increase the speed of protein crystallographic data collection at SSRL. The detector will be placed on our new MAD beamline 9-2 when that beamline has been completed sometime early next year. In the meantime, we plan to make this new CCD detector available to users on our existing MAD beamline 1-5AD from the beginning of the next running period, for a few months.
At the end of September 1997, we received our new "fast" MAR-Research imaging plate data collection system, called the MAR345. We are currently installing this new system on our recently c ompleted beamline 9-1, and plan to make it available to users after some initial tests on that line at the beginning of the next run. This new imaging plate detector has a programmable plate diameter of up to 345 mm, and a readout time that is approximately three times as fast as the MAR300 systems that have been used on both BL7-1 and BL9-1 in the past (MAR-Research kindly loaned us a second MAR300 for use on beamline 9-1 for the past year during the development of their new MAR345 detector system). The pixel size of the MAR345 can be set to either 100 µm or 150 µm, compared to 150 mm for the MAR300. To give you an example of the speed of the new MAR345 detector, the total readout and erasure time for an imaging plate diameter of 300 mm and a pixel size of 150 µm, is 66 seconds, compared to 210 seconds for the MAR300. Our new MAR345 also incorporates an extra long base unit that will allow the crystal-to-detector distance to be varied between 75 mm and 800 mm. When the maximum imaging plate diameter of 345 mm and the minimum pixel size of 100 µm is selected, each diffraction image requires 23.8 MB (uncompressed) and approximately 4 or 5 MB (compressed) of storage. For this reason, we have purchased a disk farm capable of storing ~80 GB of data on BL9-1.
The MAR300 imaging plate system, that has given us excellent service for many years on our wiggler beamline 7-1, was sent back to the factory in Hamburg after the end of the last run for a complete overhaul. The imaging plate, lasers, power supplies, etc., have been replaced, and the system has been recalibrated. It has already arrived back at SSRL, and has been reinstalled on beamline 7-1 ready for the next run.
SSRL has just taken delivery of a Huber Kappa-geometry goniometer (Kappa angle of 60 degrees) for use on our new MAD beamline 9-2 when that line has been completed. This goniometer, recently developed by Huber in collaboration with Wilfried Schildkamp (of BIOCARS at the APS and the Univ. of Chicago), incorporates a number of excellent features for protein crystallographic experiments. One of the greatest virtues of a Kappa-geometry goniometer is the ease of access to, and the unhindered space surrounding, the sample crystal. The simple counterweighted 3-axis system (without a large chi plane) provides a "sphere of uncertainty" at the center of the machine of approximately 12.5 µm radius. This will allow us to very accurately align, in any general orientation, and collect data from very small protein crystals. The sample crystal support incorporates motorized X, Y and Z translations, with enough translation along the phi axis (+/- 12.5 µm) to accommodate a large range of cryogenic mounting equipment, pressure cells, flow cells, etc. Using a CCD TV telescope, these automated alignment translations X, Y and Z will permit us to align the sample from a remote computer terminal, inside or outside the experimental enclosure. The motion of the omega, kappa and phi axes are sufficiently rapid to make it feasible to collect anomalous scattering data using the inverse beam technique. In contrast to Eulerian-geometry goniometers, the Kappa-geometry also permits a greater angular range to be covered before the goniometer hardware would interfere spatially with the input X-ray collimator or the X-ray detector. Since we are working towards the goal of providing similar experimental apparatus (and software interfaces) on each of our protein crystallography beamlines, for the convenience of our user groups and to simplify upkeep, two more of these goniometers are now on order. We plan to equip each of our new advanced detector systems with such a Kappa-geometry goniometer in the next two years.
The MAD phasing beamline 1-5AD was heavily used during the 1996-1997 run. The large majority of the users collected multi-wavelength data, principally from proteins genetically engineered to contain selenomethionine. Users have so far reported solving four structures using MAD data collected on this line this year. MAD data sets were also collected this year from other proteins for which the results are not yet available. We also collected data on three proteins that were not MAD projects but had unit cells too large to be accommodated on our other beamlines. During the first two months of the run a MAR-Research automated imaging plate detector was used on BL1-5AD. While no faster than the standard manual system used on BL1-5AD, data collection was much easier, and users had more time to evaluate their data while it was being collected.
For the bulk of the run we reverted to the use of Fuji imaging plates scanned off-line. A room near to the beamline was specially equipped for the scanner. This made scanning and erasing the imaging plates, and reloading the cassettes, considerably more convenient. In September, the computer that controls the Fuji imaging plate scanner was upgraded to a Power Macintosh.
Beamline 9-1 has been scheduled regularly for users since March 1997, and has produced a wealth of high quality data on many exciting projects. The first attempts were made to collect diffraction data to very high resolution (better than 1 Å) for which the X-ray wavelength was set to 0.77 Å. This allowed for the measurement of very complete and high quality data to a maximum resolution of 0.78 Å on a 29 kDa enzyme. The resulting electron density map is of astounding quality and reveals details not seen in proteins of that size before. Various modifications were made to beamline hardware, and data collection strategies were designed in the course of these tests. We anticipate ultra-high resolution data collection to become a routine experiment at SSRL in the near future.
The development of techniques for preparing heavy atom derivatives, to be used for the multiple isomorphous replacement (MIR) method of structure determination, has continued. An attempt at preparing a xenon derivative at cryogenic temperature proved to be successful and an apparatus to perform these experiments was designed and fabricated for general use at the crystallography beamlines. SSRL staff scientists, in collaboration with a number of visiting experimenters, were successful in producing heavy atom derivatives for a variety of proteins at cryogenic temperature. The success rate of xenon binding appears to be about 50% as observed for room temperature experiments.
|