With complete genome sequences available, it is now possible to examine all of the proteins in a genome for involvement in multisubunit assemblies. How different proteins are able to form stable complexes is of fundamental interest from the perspective of protein structure and folding. In addition, identifying proteins that physically interact can provide valuable clues about their biochemical and biological functions. Mapping domains within proteins that are responsible for oligomerization is an important part of structure-function analysis. This application describes experiments to simultaneously identify and localize oligomerization domains on a genome-wide scale. Genomic DNA fragments from S. cerevisiae that encode motifs that can self-assemble will be identified by a genetic approach based on gene fusion methods using E. coli as a host. Libraries of yeast DNA fragments cloned as gene fusions to the DNA binding domain of bacteriophage lambda cI repressor will be subjected to selection for repressor activity, which requires assembly into dimers or higher oligomers. Initial characterization of candidate motifs will exploit the unique ability of the repressor system to distinguish between dimers and higher oligomeric forms in vivo. While the selection and characterization of oligomerization domains from yeast is in progress, the search will be extended to find self-assembling domains from two bacteria, E. coli and M. tuberculosis, and two filamentous fungi, N. crassa and A. fumigatus. Although the primary focus of this proposal is on homotypic interactions, methods will be developed to use combinations of libraries in E. coli-based two-hybrid systems to examine protein motifs from S. cerevisiae that are sufficient to form heterotypic complexes. Oligomerization domains will be expressed and purified from E. coli. Size exclusion chromatography and analytical ultracentrifugation will be used to determine their oligomerization states. The boundaries of the domains that are necessary and sufficient to form stable complexes will be determined by partial proteolysis, followed by analysis of protease resistant fragments by N-terminal peptide sequencing and mass spectrometry. Structures of soluble oligomerization domains will be determined by X-ray crystallography. Expression vectors will be developed to use the oligomerization domains as """"""""dominant negative"""""""" inhibitors in S. cerevisiae and in E. coli. This work will contribute to human health by providing important insights into protein taxonomy, materials for protein design, new tools for genetic studies in model organisms (S. cerevisiae and E. coli) and important human pathogens (M. tuberculosis and A. fumigatus), and new drug targets based on protein-protein interactions.
