eIF2Bε phosphorylation activates integrated stress response-mediated protein translation and links stress signalling to amyloidogenesis in Alzheimer’s disease

A central etiopathogenic event in Alzheimer’s disease (AD) is the accumulation of amyloid β-peptide (Aβ) derived from the amyloidogenic processing of the amyloid precursor protein, a pathway initiated by BACE1. Chronic activation of the Integrated Stress Response is linked to AD, typically through eIF2α phosphorylation, which selectively enhances translation of key stress-responsive mRNAs like ATF4 and BACE1. We investigated a novel regulatory mechanism mediated by Glycogen Synthase Kinase-3β (GSK-3β) a hyperactive kinase in AD, on the translational factor eIF2B, the guanine nucleotide exchange factor (GEF) for eIF2. The methodology combined cellular and molecular biology approaches (western blot, ELISA, immunofluorescence, and luciferase assays) using pharmacological inhibitors and plasmid transfection in established cell models, along with computational structural modeling (AlphaFold3). Findings were validated in post-mortem human brain tissue, with statistical analyses (t-tests and ANOVA) applied throughout. We treated cells (SH-SY5Y) with a GSK-3β inhibitor (NP031112) or overexpressed a constitutively active GSK-3β mutant (GSK-3β-S9A) to modulate GSK-3β activity. GSK-3β inhibition significantly reduced ATF4 and BACE1 protein levels in SH-SY5Y cells in a dose-dependent manner as we tested by western blot ( P <0.001), independent of eIF2α phosphorylation. This effect was translational, as the inhibitor still reduced BACE1 levels when transcription was blocked. Crucially, NP031112 reversed stress-induced eIF2Bε phosphorylation at Ser540 ( P <0.001). Furthermore, H ₂ O ₂ -induced Aβ _1-42 secretion by SH-SY5Y cells was significantly reversed by GSK-3β inhibition ( P <0.001). Luciferase reporter assays using the BACE1 5’ untranslated region (5’ UTR) confirmed that both eIF2Bε silencing and a phosphomimetic mutant eIF2Bε-S540E increased BACE1 5’ UTR-driven translation ( P <0.001), demonstrating that reduced functional eIF2B enhances BACE1 translation. Finally, immunohistofluorescence and western blot analysis of human AD hippocampi showed a significant increase in pS540-eIF2Bε, ATF4, and BACE1 levels in AD patients compared to non-demented controls ( P <0.001 for all three). Our findings establish a novel GSK-3β-eIF2Bε-BACE1 axis that links stress response dysregulation to amyloidogenesis, independent of canonical eIF2α phosphorylation. The phosphorylation of eIF2Bε functionally mimics the effects of eIF2α phosphorylation by reducing the available functional eIF2B pool, thereby reprogramming translation to favour the production of stress proteins like ATF4 and BACE1. We proposed eIF2Bε phosphorylation is a key etiological mechanism in AD.