Producing well diffracting is one of the bottlenecks in macromolecular crystallography. Commercial sparse matrix are nowadays vastly used for the initial screening of crystallization conditions (1). The design of these screens is based on conditions that have been successful in the past with other proteins (2). However, these screens are have their weaknesses firstly because a substantial amount of structures were solved by high-throughput consortia. These aimed at the beginning to develop methods and solve the structures of procaryotically expressed, secreted and highly soluble proteins (3). Secondly, repeated use of sparse matrix screens without adding new conditions just oversamples the pre-selected conditions but will not change the parameter space to a maybe more efficient and comprehensive one.
Crystallization is a thermodynamically driven process, which depends on a multitude of parameters that are not yet completely understood. However there is evidence that the propensity of a protein to crystallize in a certain condition is correlated to its melting temperature (Tm) (4).
With the aim to maximize the probability of success, we developed a crystallization strategy, where underlying concept is to mainly employ compounds, which increase the melting temperature of the protein. In the first step, the protein is subjected to thermofluor assays, in order to determine which compounds and which pH have the major impact on its stability. In the second step, the protein is then subjected to a customized crystallization screen designed to comprise compounds that were identified to increase the Tm.
The customized strategy yielded diffraction-quality crystals of a wide range of model as well as not-crystallised in-house proteins, where the standard approaches failed. Furthermore, the inherently simple design, modular and flexible nature of the platform makes it easy to modify and optimize certain steps. Information gained through the approach can be also used to improve the sample monodispersity, stability for storage, and even derive a possible function of not yet annotated proteins. Herein we will report on the results of our customized crystallization strategy, and give an outlook on future experiments and applications.