Protein kinases are key regulatory molecules that participate in many biochemical pathways. As catalysts, these enzymes can amplify intra- or extracellular signals. By virtue of their broad yet limited substrate specificity, they are able to regulate more than one substrate protein at a time, coordinating intracellular metabolism. Understanding how protein kinases achieve the specificity necessary to regulate cell processes is an important problem in biology, and is the long-term goal of this proposal. Specifically, a detailed structural study of TPK1, the cAMP-dependent protein kinase catalytic subunit of Saccharomyces cerevisiae, is proposed. Existing crystals of TPK1 that diffract to 2.2 A resolution will be derivatized with heavy atom compounds, solving the crystallographic phase problem, and yielding its three-dimensional structure. The structure will reveal the location of key residues previously demonstrated to be involved in substrate recognition and catalysis, and reveal the overall folding of a protein kinase domain. This work will be extended by predicting the structure of other protein kinases that differ from TPK1 in substrate specificity using comparative modeling methods. The resulting prediction of protein kinase substrate selectively will be tested by in vitro mutagenesis. Chimeric kinases will be synthesized and characterized to relate the structural features predicted by computer modeling to enzymatic activity. A complete understanding of protein kinase structure will speed the development of specific inhibitors potentially useful in the treatment of neoplastic disease.