Glycogen synthase kinase 3 (GSK3) is a serine/threonine kinase that has been implicated in pathological conditions such as diabetes and Alzheimer's disease. We report the characterization of a GSK3 inhibitor, AR-A014418, which inhibits GSK3 (IC50 = 104 ± 27 nM), in an ATP-competitive manner (Ki = 38 nM). AR-A014418 does not significantly inhibit cdk2 or cdk5 (IC50 > 100 μM) or 26 other kinases demonstrating high specificity for GSK3. We report the co-crystallization of AR-A014418 with the GSK3β protein and provide a description of the interactions within the ATP pocket, as well as an understanding of the structural basis for the selectivity of AR-A014418. AR-A014418 inhibits tau phosphorylation at a GSK3-specific site (Ser-396) in cells stably expressing human four-repeat tau protein. AR-A014418 protects N2A neuroblastoma cells against cell death mediated by inhibition of the phosphatidylinositol 3-kinase/protein kinase B survival pathway. Furthermore, AR-A014418 inhibits neurodegeneration mediated by β-amyloid peptide in hippocampal slices. AR-A014418 may thus have important applications as a tool to elucidate the role of GSK3 in cellular signaling and possibly in Alzheimer's disease. AR-A014418 is the first compound of a family of specific inhibitors of GSK3 that does not significantly inhibit closely related kinases such as cdk2 or cdk5.
Tau protein hyperphosphorylation and aggregation into neurofibrillary tangles are characteristic features of several neurodegenerative disorders referred to as tauopathies. Among them, frontotemporal dementia and Parkinsonism linked to chromosome 17
A neuroblastoma protein related to the brain microtubule-associated protein, MAP-1B, as determined by immunoprecipitation and coassembly with brain microtubules, becomes phosphorylated when N2A mouse neuroblastoma cells are induced to generate microtubule-containing neurites. To characterize the protein kinases that may be involved in this in vivo phosphorylation of MAP-1B, we have studied its in vitro phosphorylation. In brain microtubule protein, MAP-1B appears to be phosphorylated in vitro by an endogenous casein kinase II-like activity which also phosphorylates the related protein MAP-1A but scarcely phosphorylates MAP-2. A similar kinase activity has been detected in cell-free extracts of differentiating N2A cells. Using brain MAP preparations devoid of endogenous kinase activities and different purified protein kinases, we have found that MAP-1B is barely phosphorylated by cAMP-dependent protein kinase, Ca/calmodulin-dependent protein kinase, or Ca/phospholipid-dependent protein kinase whereas MAP-1B is one of the preferred substrates, together with MAP-1A, for casein kinase II. Brain MAP-1B phosphorylated in vitro by casein kinase II efficiently coassembles with microtubule proteins in the same way as in vivo phosphorylated MAP-1B from neuroblastoma cells. Furthermore, the phosphopeptide patterns of brain MAP-1B phosphorylated in vitro by either purified casein kinase II or an extract obtained from differentiating neuroblastoma cells are identical to each other and similar to that of in vivo phosphorylated neuroblastoma MAP-1B. Thus, we suggest that the observed phosphorylation of a protein identified as MAP-1B during neurite outgrowth is mainly due to the activation of a casein kinase II-related activity in differentiating neuroblastoma cells. This kinase activity, previously implicated in beta-tubulin phosphorylation (Serrano, L., J. Díaz-Nido, F. Wandosell, and J. Avila, 1987. J. Cell Biol. 105: 1731-1739), may consequently have an important role in posttranslational modifications of microtubule proteins required for neuronal differentiation.