Experimental methods in structural biology: Protein crystallization and X-ray crystallography

The intention is to dedicate this chapter to the basics of the major experimental methods used in tertiary protein structure determination. One of these methods, X-ray crystallography, has made the largest contribution to our understanding of protein structures, although the other methods have complemented our data when crystallography for one or other reason could not be used.

The primary experimental methods used in the study of tertiary protein structure include:
Protein crystallography
◦ Electron microscopy (and especially Cryo-electron microscopy): Electron crystallography and single particle reconstruction
◦ Nuclear magnetic resonance (NMR spectroscopy)
◦ Small-angle X-ray and neutron scattering (SAXS and SANS)
Homology modelling

Each experimental method has its advantages and limitations. Thus, electron microscopy, and particularly Cryo-EM, which provides higher resolution, can be used to study relatively large objects, like cellular organelles or large macromolecular complexes, using the method of single-particle reconstruction. An advantage of the method of single-particle reconstruction, in comparison to protein crystallography, is that it does not require the protein to be crystallized, since
crystallization in many cases may be difficult and may require a lot of effort. However, in electron crystallography, which is primarily used for membrane proteins, we do need crystals, so called two-dimensional crystals of a protein. Cryo-EM also requires small amounts of material, which is an advantage in comparison to both crystallography and NMR spectroscopy. NMR requires much larger amounts of material and in addition, the protein under study needs to be stable at room temperature under a rather long time of data acquisition. One of the limitations of Cryo-EM is that the resolution obtained is generally limited in comparison to the resolution obtained from NMR spectroscopy or protein crystallography.

Protein crystallography may provide atomic or near atomic resolution, when small details of the protein structure can be resolved with very high accuracy. Although NMR spectroscopy does not provide this level of resolution, the method proved to be valuable when details of the dynamics of the system needed to be studied or when a protein is difficult to crystallize.
SAXS and SANS, like for example single-particle electron microscopy, provide limited resolution of the structures. Like NMR, measurements are performed in solution, thus having the advantages of controlling the conditions of the experiment directly in solution.
Homology modeling may also be used for obtaining three-dimensional structural information of protein structures. However, for accurate modeling we need high percentage identity between the amino acid sequence of the given protein and its homologue for which the tertiary structure is known (the template). Additional methods used in obtaining partial (local) structural information include mass spectrometry, analytical ultracentrifugation, various fluorescent spectroscopic methods, etc.

An outline of X-ray crystallography
Crystallography starts from a crystal and to get crystals the protein needs to be crystallized. People often say: Crystallization is an art and not a science. It is true to a certain extent, there are still some general principles which need to be followed, and most importantly, there are different methods, which have been developed to facilitate protein crystallization. After obtaining crystals it is time for the X-ray crystallographic experiment, which is placing the crystal in an X-ray beam, rotating it and collecting diffraction data. Once data have been obtained, the rest of the work will be done using specialized computer programs. To start with, I will first give an overview of the method of crystallization and crystallization tools, followed by an overview of crystallography, X-ray data collection, refinement and structure quality:
Crystallization tools
Crystallography overview
X-ray data collection
Model building
Structure quality