Crystal structure of JNK3: a kinase implicated in neuronal apoptosis.

Abstract

The c-Jun N-terminal kinases (JNKs) are members of the mitogen-activated protein (MAP) kinase family, and regulate signal transduction in response to environmental stress. Activation and nuclear localization of JNK3, a neuronal-specific isoform of JNK, has been associated with hypoxic and ischemic damage of CA1 neurons in the hippocampus. Knockout mice lacking JNK3 showed reduced apoptosis of hippocampal neurons and reduced seizure induced by kainic acid, a glutamate-receptor agonist. Thus, JNK3 may be important in the pathology of neurological disorders and is of significant medical interest. We report here the structure of unphosphorylated JNK3 in complex with adenylyl imidodiphosphate, an ATP analog. JNK3 has a typical kinase fold, with the ATP-binding site situated within a cleft between the N- and C-terminal domains. In contrast to other known MAP kinase structures, the ATP-binding site of JNK3 is well ordered; the glycine-rich nucleotide-binding sequence forms a beta-strand-turn-beta-strand structure over the nucleotide. Unphosphorylated JNK3 assumes an open conformation, in which the N- and C-terminal domains are twisted apart relative to their positions in cAMP-dependent protein kinase. The rotation leads to the misalignment of some of the catalytic residues. The phosphorylation lip of JNK3 partially blocks the substrate-binding site. This is the first JNK structure to be determined, providing a unique opportunity to compare structures from the three MAP kinase subfamilies. The structure reveals atomic-level details of the shape of JNK3 and the interactions between the kinase and the nucleotide. The misalignment of catalytic residues and occlusion of the active site by the phosphorylation lip may account for the low activity of unphosphorylated JNK3. The structure provides a framework for understanding the substrate specificity of different JNK isoforms, and should aid the design of selective JNK3 inhibitors.