W0076

The Use of Robotics in High Throughput Crystallography. M. Gary Newton¹, John P. Rose¹, Zhi-Jie Liu¹, Doowon Lee¹, Ashit Shah¹, Keith Crane², Rob Neeper³, Rhett Affleck³, Shu-Huey Chang¹, Wolfram Tempel¹ and Bi-Cheng Wang¹: ¹Southeast Collaboratory for Structural Genomics, Dept. of Biochemistry and Molecular Biology, Univ. of Georgia, Athens, GA 30602, ²Rigaku/MSC, The Woodlands, TX 77381 and ³Discovery Partners International, 9640 Towne Center Dr., San Diego, CA 92121.

Two major bottlenecks have been encountered in many high-throughput (HTP) crystallography laboratories: (1) tracking, recording and screening the enormous number of trials required to find suitable crystallization conditions and (2) efficient searching to find crystals which diffract well enough for an X-ray analysis. Our laboratory is using commercially available robotic systems to overcome these bottlenecks.

Bottleneck (1). The Discovery Partners Crystal Farm robot is a controlled environmental chamber that can store, track and manipulate up to 600 plates that are assigned unique barcodes. These plates are prepared for initial screening by a Douglas Instruments or a Genomics Solutions Honeybee SD robot. Each plate inside Crystal Farm can be selected and moved to an internal microscope stage for inspection. The image of each well can be stored in a database and viewed on a computer screen. Once a successful crystallization has been identified, the crystallization conditions are optimized using the Douglas robot. Optimized crystals are passed on to the next process described below.

Bottleneck (2). The evaluation of crystal diffraction quality is aided by an MSC/Rigaku ACTOR/DIRECTOR robot system that automatically loads, aligns, collects data and unloads loop-mounted crystals to the goniometer of an MSC/Rigaku Saturn 92 detector. During the entire process, the crystals are maintained at LN₂ temperature. The 5 magazines in the LN₂ Dewar of the ACTOR will hold a maximum of 60 crystals. Initial X-ray screening of the 60 crystals requires about 5 minutes each for a total of 5 hours without intervention.

The efficacy of these robotic systems and our experiences in using them will be reported.

Work supported in part with funds from the National Institute of Health (GM62407), The Georgia Research Alliance and the University of Georgia Research Foundation.