Featured Article - February 2011
Short description: The PAS-domain regulated kinase (PASK) has been implicated in nutrition sensing and signaling to facilitate a metabolic response to changing nutritional status. The crystal structure of PASK now indicates a basis for the unusual activation properties and substrate selectivity of this kinase.
Several kinases have been linked to sensing and responding to changing nutritional status, and these components are key to coordinating activities of the cell, such as growth and activation of relevant biochemical pathways, with metabolic demand. One such kinase is the Per-Arnt-Sim (PAS) domain–containing kinase PASK. Both mammalian and Saccharomyces cerevisiae PASK orthologs have been linked to sensing and responding to changes in metabolic status, but the substrates of mammalian PASK are unclear. Although PASK-mediated sensing is mysterious, it is likely to be a function of the PAS domain, given its role as a sensory module in light, oxygen and other sensing systems, as well as its ability to bind small molecules. The output of this sensing probably occurs through the canonical C terminal Ser–Thr kinase domain and Rutter, Burley and colleagues from the New York Structural Genomics Research Center (PSI NYSGXRC) have now solved the crystal structure of and characterized the kinase domain of human PASK.
Overall, the structure of the PASK kinase domain resembles a typical two-lobed kinase domain, with a β-sheet rich N-terminal lobe and a more α-helical C-terminal region. However, it contains an unusual β-hairpin in the N-terminal lobe region that is also found in the PIM1 protein kinase. Indeed, the PASK kinase domain shares aromatic residues with PIM1 in this region. This conservation, and the conserved nature of a number of residues in this region with those in other PASK orthologs, suggests that this unusual structural feature may be linked to regulation, an idea that merits further follow up.
PASK is also unusual in that the authors report that phosphorylation of the activation loop in the kinase is not required for function. Although most PASK orthologs retain a phosphorylatable residue—threonine—in this loop, mutation to alanine or a phosphomimetic residue does not alter kinase activity on an in vitro substrate or autocatalytic activity in vivo. Phosphorylation of this loop is thought to be a requirement for activity of many protein kinases, as it helps to promote interactions between the N and C lobes of the kinase, thus stabilizing the active conformation and substrate-binding interface. The structure does elucidate interactions that might compensate for the lack of a requirement for activation-loop phosphorylation. For example, in many eukaryotic protein kinases, a key lysine is thought to be involved in charge repulsion of a conserved so-called HRD loop in the inactive state, with phosphorylation of the activation loop then neutralizing the positive charge of the lysine. In both PIM1 and PASK, this lysine is altered to an alanine, suggesting that phosphorylation would not be required for neutralization in these cases. This is also true of another kinase that does not require phosphorylation of this activation loop for activity, CHK1, which also lacks a lysine and instead carries a threonine at this position. These and other observations suggest how it is that PASK activity does not depend on activation-loop phosphorylation, addressing one aspect of activity regulation, but the authors also go on to examine another mysterious aspect of PASK function, substrate specificity.
By screening a combinatorial peptide library, the authors take a first step toward examining this last question. On the basis of this analysis, PASK shares a preference for basic residues, and in particular arginine, with the CAMK kinase group at residue −3 relative to the phosphorylation site. Unusually however, it is selective for either histidine or arginine at the −5 position. In short, the human PASK structure suggests a basis for the unusual activity of this kinase, as well as elucidating its substrate preferences. This may facilitate an understanding of the regulation and the targets of this kinase, and thus may ultimately lead to a deeper understanding of metabolic sensing and the response to changes in the nutritional status of a cell.
C.K. Kikani et al. Structural basis of PAS domain-regulated kinase (PASK) activation in the absence of activation loop phosphorylation.
J. Biol. Chem 285, 41034-41042 (2010). doi:10.1074/jbc.M110.157594