X-ray crystallography is a method of seeing the structure of molecules that can’t be seen with a microscope. To do this, first the molecule of interest, such as a protein, must be arranged in crystal form. This is accomplished by putting the protein into a solution that encourages the molecules to align themselves in an ordered fashion. This solution usually has a high salt or polymer concentration and often is pH regulated to keep the protein folded in a physiologically relevant way. Care must be taken not to over concentrate this solution or the protein will precipitate out of solution in a disordered form which can not be analyzed. The crystallization step is often the most time consuming and some proteins are not amenable to crystallization, making this the rate limiting step in crystallography.
Next, the crystal is mounted in an x-ray diffractometer, a machine that can shoot a narrow band of x-rays through the crystal and catch their change in location as they come out the other side. The x-rays pass through the crystal until they hit an atom, or more accurately, the electric field generated by the atom's electrons. The electric field deflects (diffracts) the x-rays in a specific manner.
Because in a crystal the molecules are arranged in a regular pattern the light waves can constructively interfere and enhance the signal of the diffracted x-rays sufficiently for detection. The reflections are captured and transferred to a computer.
Because proteins are three dimensional, a single two dimensional image cannot give sufficient information to reconstruct what the entire protein looks like, just as a photograph of part of a person’s head cannot tell you how they look when viewed from all other angles. For this reason, images of the protein crystal from all angles are needed to get a complete picture. Once all images have been taken a computer assembles the spots from all the images and pieces together a model of the three dimensional protein. Because the x-rays have been deflected by the electron fields of the atoms the result is a picture of where the electrons are in the protein.
By using the shape of the electron density and the known amino acid sequence of the protein, it is possible to create an accurate model of the position of each atom and bond within the protein.
The final protein model can be presented in a variety of ways.
The nature of a crystal requires that the shape of the molecules within it be identical. This presents a problem when determining the structure of a protein that experiences dramatic conformational change as part of its function. The crystal structure obtained is a snapshot of the protein at one point and in order to obtain a more complete picture of a protein many different crystals of the same protein may be grown in order to obtain a model of different conformations of the same protein.
Virtual Screening
Traditional drug screening involves physically testing a large library of compounds (thousands or millions) against a target of interest and then using the 'hits', the compounds that show desired activity, as a template to produce better compounds. This process is expensive and time consuming, and is consequently only available to large organizations capable of funding such research. Virtual Screening (VS) allows a much larger number of researchers the ability to screen libraries of compounds without the expense of a physical compound library.
Compounds are tested in silico, that is, fit into the active site of the enzyme target by a computer according to criteria set by the researcher. By using existing knowledge about the target and it's interactions and other known physical properties of the target, VS allows researchers to pre-select compounds which show some potential interaction with a target. This saves time and decreases the cost associated with developing a new compound.
Structure Based Drug Design
Structure Based Drug Design (SBDD) is a process of developing drugs against a target based on the structure of the target. Traditional high through-put methods of drug development are based on luck - screening a library of random compounds in the hopes of finding a compound that acts against the target. This method is expensive and has a very low success rate. However, it is possible to tailor a potential drug to a target by using the wealth of structural information currently available, this is the basis for SBDD.
Many drugs are enzyme inhibitors which bind the protein's active site and inhibit the natural reaction from occurring. By recognizing the interactions in the active site with the natural substrate or any known inhibitors, it is possible to design a new compound which can inhibit the activity of the enzyme. As more structural information becomes available, our understanding of the chemical interplay within molecules becomes more refined and allows us to better use this technique.
The role of x-ray crystallography in SBDD is to allow us to directly observe the interaction between the compound of interest and the target, by interacting the two and obtaining the structure. This process is iterative. The information from this structure can aid in making new changes to the drug and then viewing the interaction between the new compound and the target. This can greatly increase the efficiency in making new and effective drugs by reducing the element of chance in drug design.
There are several commercially available drugs that have been designed using SBDD and more are in development. Tamiflu (influenza), Gleevec (leukemia) and Crixivan (AIDS) are all examples of drugs designed using structural information. Future drug development will increasingly depend on the knowledge and the design of new compounds instead of relying on luck to find a useful drug among compounds that already exist.
