© 2001, National Academy of Sciences, U.S.A.

Protein phosphorylation is tightly regulated in space and time but it has so far been almost impossible to study this aspect of cell signalling. Roger Tsien and colleagues have now developed a new class of genetically encoded fluorescent reporters that might allow us to follow the activation of virtually any kinase in living cells.

The new probes contain a substrate domain specific for a given protein kinase and a phosphorylation recognition domain that binds the phosphorylated substrate domain. This construct is sandwiched between two green-fluorescent-protein (GFP) variants that are compatible for FRET (fluorescence resonance energy transfer), for example cyan fluorescent protein and yellow fluorescent protein. If the substrate domain is phosphorylated on activation of the kinase, it undergoes an intramolecular interaction with the phosphorylation recognition domain, which induces a conformational change that alters the distance and/or the relative orientation between the two fluorescent proteins and hence the efficiency of FRET. In principle, the FRET change is reversible if a phosphatase dephosphorylates the substrate.

The authors used various chimaeras built according to this blueprint to follow the activity of four well known kinases: the serine/threonine kinase PKA (protein kinase A) and the tyrosine kinases Src, Abl and EGFR (epidermal growth factor receptor). They transfected cells with the constructs and then measured FRET before and after stimulation of kinase activity.

Adding growth factors to cells to stimulate individual tyrosine kinases caused, within minutes, a readily measurable 25–35% change in FRET. In the case of Src and EGFR, the signal started at the plasma membrane and then spread towards the cell centre. Abl, on the other hand was most active in membrane ruffles, consistent with its known function in cell migration.

Activation of PKA by forskolin, also caused a rapid 25–50% change in FRET, and, in this case, the kinase was active throughout the cytoplasm. However, when the reporter was relocalized to the nucleus by addition of a nuclear localization signal, the FRET change was much delayed, whereas tethering the reporter to PKA accelerated the response. This shows how important substrate localization is for susceptibility to PKA in vivo.

The new probes show a lot of promise, and optimized versions — in particular with respect to specificity — will certainly become invaluable tools for studying targeting and compartmentation of signalling in living cells.