Serendipitous crystallization of cyanase from Serratia : structure determination and a hypothesis for this recurring observation

The serendipitous crystallization of a protein other than the target of a study is not uncommon and often leads to a diffraction dataset that cannot be solved. Software tools such as SIMBAD from the CCP4 suite overcome this challenge by performing a search with common contaminants and by searching the PDB for deposited protein structures that crystallized with similar unit cells. Once solved, however, a remaining question is the original source of the protein that mistakenly formed the crystal. The least surprising situation occurs if the protein contaminant derives from the heterologous expression system. If a protein that is expressed in Escherichia coli is not adequately purified, or if the identity of a protein is confirmed only through gel electrophoresis and not through activity or peptide-mapping techniques, one may simply crystallize the wrong protein. The presence of a contaminant from a species other than the expression system is a little harder to explain. In one recurring example, crystals of a cyanase enzyme from Serratia species have been reported multiple times. Here, I report yet another example of the crystallization of a cyanase enzyme and the fifteen-year journey to solve the 2.4 Å resolution structure of this decameric protein. I discuss this repeated observation of serendipitous crystallization of this specific enzyme and propose the hypothesis that azide, a common antimicrobial added to biochemical buffers to prevent microbial growth, induces the production of cyanase by contaminating bacteria within the crystallization experiment, leading to protein accumulation and serendipitous crystallization.