Single mutations to tyrosine or glutamate improve the crystallizability and crystal diffraction properties of a flexible two-domain protein

This case report describes single surface substitutions that improve the crystallizability and diffraction properties of a flexible two-domain protein. InlB₃₉₂ comprises the internalin domain and the B repeat of the Listeria monocytogenes invasion protein InlB. The InlB₃₉₂ wild type yielded very few poorly reproducible hits in crystallization screens and the crystals had a diffraction limit of worse than 3.0 Å. It seems reasonable to assume that this crystallization bottleneck is caused by interdomain flexibility, given that crystals of the isolated internalin domain or B repeat diffract to high resolution. A previously identified variant, T332E, showed improved crystallization and diffraction. Here, two additional InlB₃₉₂ variants are described with single threonine-to-tyrosine or valine-to-glutamate substitutions that produced crystals directly in initial screens and, without optimization, diffracted to 1.6 and 1.45 Å resolution, respectively. The mutated residues do not participate in intramolecular interdomain interactions but mediate crystal contacts, indicating that specific surface properties, rather than interdomain flexibility per se, impede the crystallization of wild-type InlB₃₉₂. Notably, the beneficial glutamate substitutions contrast with the generally recognized underrepresentation of glutamate in crystal contacts and the high entropic cost of fixing an otherwise flexible side chain with many rotatable bonds in a crystal contact. The reported results suggest that surface mutations can help crystallization even if they increase the entropy of the respective residue. More broadly, the observations are consistent with the hypothesis that negative evolutionary design limits fortuitous lattice formation of proteins and the resulting expectation that random mutations of surface residues are likely to improve crystallizability.