This grant application addresses the major obstacle to using crystallographic methods to gain insight into biological function, which is the failure of most naturally occurring proteins to yield crystals suitable for x-ray structure determination. The goal of the current project is to develop methods for rational engineering of the sequence of a protein to produce high quality crystals. Structural genomics consortia systematically confirmed that crystallization is the major obstacle to determining the atomic structure of proteins using x-ray diffraction methods. Previously published work from the Hunt laboratory employed computational analysis of large-scale crystallization trials to demonstrate that protein surface properties, particularly the mean entropy of exposed sidechains, are a dominant determinant of crystallization propensity. This study identified a variety of sequence properties that correlate with crystallization success, including the content of several individual amino acids. However, every one of the amino acids that positively correlates with crystallization success negatively correlates with protein solubility, and vice-versa. This effect severely limits the efficacy of using single amino acid substitutions to engineer improved protein crystallization properties because crystallization probability is low unless the initial protein preparation is monodisperse and soluble. In this application, we propose to use a suite of computational methods to identify more complex sequence epitopes that promote successful protein crystallization without impairing solubility. Computational analyses will be used to select sites for introduction of such epitopes while preserving function and stability. These novel crystallization-engineering methods will be critically evaluated and optimized using studies in which the thermodynamic stability, solubility, and crystallization properties of purified mutant proteins are determined experimentally. Our preliminary data for these proposed studies support the efficacy of the approach while also showing that the crystallization propensity of a protein is not directly coupled to its thermodynamic solubility. Therefore, if the underlying stereochemical and thermodynamics mechanisms were sufficiently well understood, then it should be possible to engineer improved protein solubility in parallel with improved crystallization propensity. The twin objectives of the research proposed in this application are to elucidate these mechanisms while also generating rigorously validated methods for improving protein crystallization and solubility. Successful development of efficient methods for engineering improved protein crystallization would facilitate a wide variety of structural/functional biology projects, including projects focused on drug-discovery and structural characterization of macromolecular complexes.
. Protein x-ray crystal structures provide unparalleled insight into the chemical properties of proteins and are widely used in drug-discovery projects, in addition to other areas of basic and applied research. However, most proteins do not crystallize well, which precludes structure determination. This application will support systematic exploration of a novel method to engineer improved protein crystallization, including rigorous studies of the related thermodynamic effects controlling solubility, which is frequently an obstacle to industrial and medical uses of proteins.
