Using two different mouse models, cells cultured from them, and also postmortem brain tissue from patients with AD, the scientists experimentally manipulated eIF2α phosphorylation in various ways, and then correlated its state with levels of BACE1 protein, Aβ generation, and amyloid plaque. Most data in this paper are already summarized on Alzforum following a talk at the 2008 Keystone conference.

New in the paper is the identification by O’Connor et al. of the operative kinase behind eIF2α’s change in behavior. It is none other than PKR-like ER kinase, or PERK for short. PERK is one of four known eIF2α kinases. ER stress, along with the unfolded protein response connected to it, as well as amino acid deprivation, are already well known to activate this kinase; indeed, PERK’s role in the unfolded protein response has its own established literature. To identify which kinase was at play in their model system, initially in vitro, the scientists used both genetics (i.e., kinase-dead and dominant-negative versions of the candidate enzyme genes) and pharmacology (i.e., molecular inhibitors). The scientists still need to verify whether PERK excessively phosphorylates eIF2α in the brain of AD models generally, or in people with AD or amnestic MCI.—Gabrielle Strobel.