Most proteins and many biological macromolecules differ from "small" molecules because the environment in which they function is aqueous. Therefore most biological macromolecules can be prompted to form crystals when the solution in which they are dissolved becomes supersaturated. The manner in which this occurs is typical of many other compounds that crystallize from solution. Membrane proteins are crystallized after being solubilized in detergent solutions and certain seed proteins like crambin that is not soluble in water, but in ethanol are crystallized by adding water instead. Apart, from the difference in solvent, standard protein crystallization and seeding techniques like streak seeding are still used in protein crystallization from organic solvent. As with other compounds, protein crystallization occurs more readily when the protein has been purified. Most proteins structures currently available are from proteins that have been crystallized from aqueous solutions. When forming a crystal individual protein molecules align themselves in a repeating series of unit cells by adopting a consistent orientation. The crystalline lattice that forms is held together by noncovalent interactions. The importance of protein crystallization is that it serves as the basis for X-ray crystallography, wherein a crystallized protein is used to determine the protein’s three-dimensional structure via X-ray diffraction. In 1934, John Desmond Bernal and his student Dorothy Hodgkin discovered that protein crystals surrounded by their mother liquor gave better diffraction patterns than dried crystals. Using pepsin, they were the first to discern the diffraction pattern of a wet, globular protein. Prior to Bernal and Hodgkin, protein crystallography had only been performed in dry conditions with inconsistent and unreliable results.
For a better part of the 20th century, progress in determining protein structure was slow due to the difficulty inherent in crystallizing proteins. When the Protein Data Bank was founded in 1971, it contained only seven structures. Since then, the pace at which protein structures are being discovered has grown exponentially, with the PDB surpassing 20,000 structures in 2003, and containing over 77,000 as of 2011.
The goal of crystallization is to produce a well-ordered crystal that is lacking in contaminants while still large enough to provide a diffraction pattern when exposed to X-rays. This diffraction pattern can then be analyzed to discern the protein’s tertiary structure. Protein crystallization is inherently difficult because of the fragile nature of protein crystals. Proteins have irregularly shaped surfaces, which results in the formation of large channels within any protein crystal. Therefore, the noncovalent bonds that hold together the lattice must often be formed through several layers of solvent molecules.
In addition to overcoming the inherent fragility of protein crystals, a number of environmental factors must also be overcome. Due to the molecular variations between individual proteins, conditions unique to each protein must be obtained for a successful crystallization. Therefore, attempting to crystallize a protein without a proven protocol can be very challenging and time consuming.
Several conditions come into factor if a protein sample will crystallize or not. Some of these factors include protein purity, pH, concentration of protein, temperature, precipitants and additives. The more homogenous a protein in solution, the better the chances are for it to form a crystal. Typical standards have the protein solution being at least 97% pure. pH conditions are very important due to the fact that different pHs can result in different packing orientations. Buffers, such as Tris-HCl, are often necessary for the maintenance of a particular pH. Precipitants, such as ammonium sulfate or polyethylene glycol, are compounds that cause the protein to precipitate out of solution.
Vapor diffusion 
Two of the most commonly used methods for protein crystallization fall under the category of vapor diffusion. These are known as the hanging drop and sitting drop methods. Both entail a droplet containing purified protein, buffer, and precipitant being allowed to equilibrate with a larger reservoir containing similar buffers and precipitants in higher concentrations. Initially, the droplet of protein solution contains an insufficient concentration of precipitant for crystallization, but as water vaporizes from the drop and transfers to the reservoir, the precipitant concentration increases to a level optimal for crystallization. Since the system is in equilibrium, these optimum conditions are maintained until the crystallization is complete.
Simply put, the hanging drop method differs from the sitting drop method in the vertical orientation of the protein solution drop within the system. It is important to mention that both methods require a closed system, that is, the system must be sealed off from the outside using an airtight container or high-vacuum grease between glass surfaces. The images to the right depict the hanging drop and sitting drop systems.
High through-put methods 
High through-put methods exist to help streamline the large number of experiments required to explore the various conditions that are necessary for successful crystal growth. There are numerous commercials kits available for order which apply preassembled ingredients in systems guaranteed to produce successful crystallization. Using such a kit, a scientist avoids the hassle of purifying a protein and determining the appropriate crystallization conditions.
Robots can be used to set up and automate large number of crystallization experiments simultaneously. What would otherwise be slow and potentially error-prone process carried out by a human can be accomplished efficiently and accurately with an automated system. Robotic crystallization systems use the same components described above, but carry out each step of the procedure quickly and with a large number of replicates. Each experiment utilizes tiny amounts of solution, and the advantage of the smaller size is two-fold: the smaller sample sizes not only cut-down on expenditure of purified protein, but smaller amounts of solution lead to quicker crystallizations. Each experiment is monitored by a camera which detects crystal growth.
Some proteins do not fold properly outside their native environment, e.g. proteins which are part of the cell membrane like ion channels and G-protein coupled receptors, their structure is altered by interacting proteins or switch between different states. All those conditions prevent crystal growth or give crystal structures which do not represent the natural structure of the protein. In order to determine the 3D structure of proteins which are hard to crystallize researchers may use nuclear magnetic resonance, also known as protein NMR, which is best suited to small proteins, or transmission electron microscopy, which is best suited to large proteins or protein complexes.
See also 
This page was reproduced (with modifications) with expressed consent from Dr. A. Malcolm Campbell. As of 2010, the original page can be found at http://www.bio.davidson.edu/Courses/Molbio/MolStudents/spring2003/Kogoy/protein.html .
- Teeter MM, Hendrickson WA (1979). "Highly ordered crystals of the plant seed protein crambin". J Mol Biol. 127 (2): 219–23. doi:10.1016/0022-2836(79)90242-0. PMID 430565.
- Stura EA, Wilson IA. (1992). In Ducruix A Giege R. Crystallization of Nucleic Acids and Proteins. New York: Oxford University Press. pp. 112–113.
- Yamano A, Heo NH, Teeter MM. (1997). "Crystal structure of Ser-22/Ile-25 form crambin confirms solvent, side chain substate correlations". J Biol Chem. 272 (15): 9597–600. doi:10.1074/jbc.272.15.9597. PMID 90924825.
- Gale R (1993). Crystallography Made Crystal Clear. Academic Press. ISBN 978-0-12-587073-3. Unknown parameter
- Berman H, Westbrook J, Feng Z, Gilliland G, Bhat T, Weissig H, Shindyalov I, Bourne P (2000). "The Protein Data Bank". Nucleic Acids Research 28 (1): 235–242. doi:10.1093/nar/28.1.235. PMC 102472. PMID 10592235.
- Branden C, Tooze J (1999). Introduction to Protein Structure. New York: Garland. pp. 374–376. ISBN 9780815303442.
- "The Crystal Robot". December 2000. Retrieved 2003-02-18.
- McRee, D (1993). Practical Protein Crystallography. San Diego: Academic Press. pp. 1–23. ISBN 978-0-12-486052-0.
- "Protein Crystallization and Dumb Luck". An essay on the haphazard side of protein crystallization by Bob Cudney: http://www.rigaku.com/downloads/journal/Vol16.2.1999/cudney.pdf
- Owens, Ray. "Protein Crystals". Backstage Science. Brady Haran.