Phosphorylation is a ubiquitous protein modification important for regulating nearly every aspect of cellular biology. Protein kinases are highly conserved and constitute one of the largest gene families. Identifying the substrates of a kinase is essential for understanding its cellular role, but doing so remains a difficult task. We have developed a high-throughput method to identify substrates of yeast protein kinases that employs a collection of yeast strains each expressing a single epitope-tagged protein and a chemical genetic strategy that permits kinase reactions to be performed in native, whole-cell extracts. Using this method, we screened 4,250 strains expressing epitope-tagged proteins and identified 24 candidate substrates of the Pho85-Pcl1 cyclin-dependent kinase, including the known substrate Rvs167. The power of this method to identify true kinase substrates is strongly supported by functional overlap and colocalization of candidate substrates and the kinase, as well as by the specificity of Pho85-Pcl1 for some of the substrates compared with another Pho85-cyclin kinase complex. This method is readily adaptable to other yeast kinases.cell signaling ͉ phosphorylation ͉ cyclin-dependent kinase T he general criteria for establishing that a protein is a substrate of a given kinase are the ability of the kinase to phosphorylate the substrate in vitro and the dependence on the activity of the kinase for phosphorylation of the substrate in vivo (1). In vivo validation of kinase substrates continues to be laborious and time-consuming. Thus, it is crucial to be able to efficiently identify candidates before committing to this step. Indirect data, such as genetic and physical interactions, can provide insights into potential substrates, but many interacting proteins and genes are not substrates (1, 2). More direct approaches using in vitro kinase reactions have been attempted in many permutations (1, 2). Purified component reactions directly measure the ability of a kinase to phosphorylate a particular substrate and can be scaled for high-throughput formats, but such conditions often compromise reaction specificity and produce false-positive results (1, 2).We sought to improve on these techniques by carrying out a biochemical screen in an environment that more closely resembles the in vivo state. To do so, we carried out kinase reactions with near physiological levels of exogenous kinase in native whole-cell extracts. These extracts maintain protein-protein interactions that may affect substrate presentation and preserve a nearly full complement of cellular proteins, including potential adaptor proteins and cofactors that could affect the reaction. These reaction conditions also preserve the native relative protein abundances and a natural competition among substrates for limited kinase.A challenge of carrying out protein kinase reactions in wholecell lysates is that one cannot attribute phosphorylation of a protein to a particular kinase in the reaction, because all kinases in the extract are capable of cat...