Crystallography

Farhad Forouhar, Ph.D., Director


X-Ray crystallography is the technique used to obtain the three dimensional structure of a macromolecule from a crystal. The macromolecule structure can provide detailed information on the site and specificity of protein-ligand interactions, protein-protein complexes, protein-nucleic-acid complexes, immune complexes, and host-pathogen (i.e. virus, bacterium) interactions. The availability of structural information can bring focus to research in the elucidation of disease mechanisms and in the development of effective drugs. Dr. Forouhar works closely with researchers to develop a project and guide them through the preparation of high quality purified protein samples suitable for crystallization. Once enough pure protein is available, optimum conditions are found to produce diffraction-quality crystals which are then subjected to an X-ray beam at a synchrotron facility. Structural features of the protein emerge as the data are processed and the structure is refined.
Users of the Crystallography Shared Resource will first meet with Dr. Forouhar for a free consultation to discuss the feasibility and strategy for obtaining a crystal structure from their protein.

Services Include:

  • Structure Modeling: Computer modeling can be performed on a protein which has a structure that is at least 30% identical to that of a protein structure deposited in the Protein Data Bank (PDB).
  • Crystal Screening: Your protein will be screened against 1536 reagents using a sophisticated robot. We need 460 – 500 µL of the pure protein for the extensive screening. Three types of images, color, UV-TPEF, and SHG will be taken on each crystal hit and analyzed for protein crystal hits versus salt crystals. These three types of images are complementary and confirmatory in distinguishing a real crystal hit (protein crystal) from a false positive hit (salt crystal). At this stage, we can usually tell whether these crystal hits are poor or promising for production of diffraction quality crystals that could lead to structure determination. In case of poor crystal hits, we will take different strategies for improving the hits from poor to diffraction quality crystals. These strategies include 1) surface mutation of charged residues, 2) screening against different protein buffers and 3) screening against small molecules that could play key roles in improving crystal quality.
  • Crystallization: Once optimum conditions for crystallization are found, those conditions will be reproduced in the lab to form crystals of sufficient size and quantity to yield high quality diffraction data.
  • Data Collection: Crystals will be shipped to an American synchrotron and X-ray data will be collected remotely.
  • Structure Determination and Refinement: The time required to determine the structure of a protein and to refine that structure is highly dependent on the size of the protein and the quality of the diffraction data. An easily completed crystal structure, such as 200-300 amino acids diffracting to 2Å resolution, might take 20 hours. A larger human protein diffracting at 3Å resolution might take many times longer. Throughout the process, we will discuss every step with the researcher.

Projected Costs
The complete fee schedule can be found here. A rough estimate of the total cost for an easy and successful crystal structure of a protein comprising 200-300 amino acids at 2Å resolution is $5000. However, most human proteins which are implicated in cancer contain flexible regions and are therefore not readily amenable to crystallization and structural studies. In such cases, extensive effort will be spent to obtaining a crystal structure of a human protein. Consequently, the total cost of delivering a protein structure to you will normally be a lot more than $5,000. Since structural studies of human proteins are case by case, we will do our best to realistically estimate the total cost of delivering a protein structure to you after we thoroughly evaluate the biophysical/biochemical properties of your protein of interest.

If there is at least one protein structure in Protein Data Bank (PDB), which shares 40% sequence identity with the protein of interest in its native wild-type form, then we will use the molecular replacement method to determine the structure of the protein of interest using the available protein model in PDB. Again we will charge you for every hour spent on performing structural studies on your protein of interest.

If there is no model for the protein of interest, then the best method for determining its structure is by using a seleno-methionine-labeled protein. In this case, the interested party will provide us with the seleno-methionine-labeled recombinant protein at the start of the service.