Preparing well-diffracting crystals is the key step in biomacromolecular crystallography and is in particular a challenge for high-throughput structure determination projects of proteins as part of structural genomics and medicinal protein crystallography projects. Another challenge is to freeze the crystals obtained to 100K in such a manner that they remain well-diffracting and with no ice crystals. This is a critical step for X-ray data collection at synchrotron beam lines. The next critical step is to place the crystal in the X-ray beam and center the crystal precisely at the proper position where cryo-stream, X-rays and rotation axes of the goniometer intersect.
The aim of our proposal is to obtain a completely automated procedure for all these steps: no manual intervention for crystal growth, cryofreezing and crystal centering. This unique approach can significantly remove all bottlenecks between protein production and the initiation of X-ray data collection for biomacromolecular crystallography. The equipment to be developed would be able to work with very small amounts of protein samples since volumes per experiment are in the low nanoliter range.
The aim of the proposed project is to develop, design, and build a complete prototype sample processor that automates all process steps from initial protein sample, through automated detection of crystal growth, though delivery of a completely characterized, cryocooled sample to the synchrotron or other x-ray diffraction facility. Specific performance goals of this prototype system include: 1) A complete pipeline: sample preparation, sample sealing, conditioned storage for crystal growth, automated identification of crystal growth, cryocooling, and characterization of the samples in anticipation of x-ray crystallography. 2) Fully-automated, """"""""hands-free"""""""" processing of samples. 3) Throughput of at least 500 samples/hour. 4) In-line image acquisition and processing for automated identification of crystal growth. The proposal has not only potential for a major impact of full automation of all steps between purified protein production and starting the X-ray data collection process in structural genomics projects. In addition, it has the potential to be coupled with combinatorial libraries of chemical compounds, which would allow thousands of compounds to be tested for crystal growth of a drug target protein in the presence of numerous different compounds.

Agency
National Institute of Health (NIH)
Institute
National Institute of General Medical Sciences (NIGMS)
Type
Research Project (R01)
Project #
5R01GM068878-02
Application #
6798177
Study Section
Special Emphasis Panel (ZRG1-SSS-6 (10))
Program Officer
Edmonds, Charles G
Project Start
2003-09-01
Project End
2006-08-31
Budget Start
2004-09-01
Budget End
2005-08-31
Support Year
2
Fiscal Year
2004
Total Cost
$343,587
Indirect Cost
Name
University of Washington
Department
Engineering (All Types)
Type
Schools of Engineering
DUNS #
605799469
City
Seattle
State
WA
Country
United States
Zip Code
98195
Pan, Shen; Shavit, Gidon; Penas-Centeno, Marta et al. (2006) Automated classification of protein crystallization images using support vector machines with scale-invariant texture and Gabor features. Acta Crystallogr D Biol Crystallogr 62:271-9