Michael Stark

An Introduction to X-Ray Crystallography

Researchers in the Strong Lab use a variety of biochemical techniques to produce immunologically significant proteins with the eventual goal of determining their three-dimensional structures, typically through X-ray crystallography.

Determining the structure of these proteins allows us to make inferences about what will bind and interact with that protein as well as how those proteins fundamentally ‘work’. For example, if we wish to produce a cancer therapy that selectively binds to cancerous cells and minimizes off-target effects, we must first know what will bind to that cancer cell, and hopefully not bind to the healthy cells of the patient. To do this, structural biologists must determine exactly what those surface proteins look like.

Despite recent advances in microscope technology, even the most powerful light microscopes fail to obtain anywhere near the resolution needed to image something as small as a single protein. To the uninitiated, a protein is simply a chain made up of any combination of the 20 different amino acid molecules. Because any of these 20 amino acids can bond with one another, a protein can look and act extremely differently from another protein. Additionally, because of the extreme diversity of these amino acid sequences, predicting how a protein will fold into shape is immensely difficult. While computational methods like AlphaFold have radically increased the accuracy of predictions, modelling large, complex, multi-unit proteins still remains a challenge computationally. Because of this, if one requires an extremely high resolution depiction of a protein, they still must turn to X-ray crystallography.

In X-ray crystallography, one prepares a highly-ordered material which consists of repeating, identical copies of the protein of interest. This crystalline solid, when placed in an X-ray beam, will then diffract those incoming X-rays in a specific way, depending on the location of the atoms in that solid. The location of those diffracted X-rays can then be used, with great precision, to determine the structure of the protein in that diffracting crystal. 

My Work as an X-Ray Crystallographer

Before one can begin an X-ray crystallography experiment one first needs some crystals to ‘shoot’. This all begins by obtaining a very pure and clean sample of one’s protein of interest. This protein is then placed in a compatible buffer, where the protein can be fully solvated, forming bonds between the protein and the solvent. From that point we can begin to attempt to coerce the protein into a crystalline configuration. However, this is often where the trouble begins.

If you ask any protein crystallographer about their work, most likely the first thing they will tell you is that it is a bit of a dark art. Or, the closest thing you can get to being a professional gambler while working in a lab, as one researcher described it to me. The reason for this is primarily because proteins require quite a bit of persuading to get them to crystallize. This involves getting the protein to forgo its favorable interactions between water molecules in the buffer solution and in exchange form billions of identical, repeating, contacts with one another. It is this formation of identical contacts that makes a crystal. 

Each protein will crystallize best in vastly different environments, requiring a crystallographer to begin each experiment by attempting hundreds of different ‘conditions’ - what we call a specific mixture of different salts, buffers, and precipitating agents. Once the protein has gone through its initial screening, a crystallographer can begin to deduce some trends as to which conditions are helping the protein to crystallize. From that point, we attempt to grow larger, singular crystals, which are hopefully robust enough to be carefully picked up and placed in the X-ray beam path.