This proposal presents a new approach to produce fluorescent biosensors, capable of reporting the spatio-temporal dynamics of endogenous protein activation in living cells. The biosensors consist of an 'affinity reagent'which binds specifically to the activated form of the target protein, coupled to a novel dye designed for live cell imaging. The dye undergoes a fluorescence change suitable for ratio imaging when the affinity reagent finds and binds the activated target. Libraries of affinity reagents will be displayed on the surface of phage, enabling selection of biosensors for specific targets using high throughput screening. This approach can produce biosensors for previously inaccessible targets, because it does not rely on identifying naturally occurring protein domains that bind activated target, or known target substrates. The approach is less perturbing than other current biosensor designs both because endogenous proteins can be examined, and because the bright dyes are directly excited for enhanced sensitivity. Phage display can be used to 'fine tune'the affinity and reversibility of the biosensor. Focused libraries will be produced to target a particular type of protein target, thus improving the efficiency of screening and the binding characteristics of the biosensor. Through computation and protein modeling, variable regions will be introduced into naturally occurring domains already targeted to phosphoproteins, and protein scaffolds will be engineered for use as biosensors. Computation and protein modeling will be used to improve the intracellular stability, labeling and expression of the sensors, and to restrict variability to the most productive regions of the structure. We will target two broadly relevant mechanisms of signaling regulation: phosphorylation and intramolecular interaction of autoinhibitory and kinase domains. Biosensors for Src, PAK, JNK, and ERK2 will be targeted as examples of such regulation, and because biosensors of these molecules will enable us to address an important biological question, for a 'real world'test of this new class of biosensor reagents. PAK, JNK, and ERK2 are each on a different, parallel pathway downstream of Src. The spatio-temporal regulation of this signaling network to produce different cellular responses will be examined.
