A mixture of fortunes: the curious determination of the structure of Escherichia coli BL21 Gab protein.

In protein crystallography, monodisperse protein samples of high purity are usually required in order to obtain diffraction-quality crystals. Here, crystals were reproducibly grown from a protein sample before its homogeneity had been determined. The sample was obtained after the first attempt to purify a recombinant target protein from an Escherichia coli cell lysate. Subsequent analysis revealed that it was a mixture of about 50 different proteins with no predominant species. Diffraction data were collected to 2.1 A and the space group was identified as I422. A molecular-replacement search with models of the expected target did not give a solution, which suggested that a contaminating E. coli protein had been crystallized. A PDB search revealed 256 structures determined in space group I422, of which 14 are E. coli proteins and two have unit-cell parameters similar to those observed. Molecular replacement with these structures showed a clear solution for one of them, the Gab protein. The structure is presented and compared with the deposited structure, from which it shows small but significant differences. The refined model contains bicine and sulfate as bound ligands, which provide insights into possible substrate-binding sites.