Abstract The ER stress sensor PERK usually serves to attenuate protein synthesis and activate apoptotic pathways. Thus, most of the work linking PERK to Alzheimer's disease and to tauophathies focuses on manipulating this activity. However, the genetic variants of PERK associated with increased risk for neurodegeneration have recently been shown to have reduced kinase activity, suggesting that the PERK pathway also has important adaptive functions that may be neuro-protective. Manipulating these protective functions is equally important for neurodegeneration. We discovered one such function of PERK, which underlies the proper localization of growth factors to axons and dendrites, which in turn is a pre-requisite for their local secretion and neuroprotection. PERK is involved in growth factor localization because its deletion in either mouse cortical neurons or in worm chemosensory neurons abolishes the correct axonal/dendritic distribution and inhibits the secretion of IGFs and TGFβ. This proposal explores how PERK controls axonal/dendritic localization and secretion of these neuroprotective growth factors. In Aim 1 we will determine if the kinase activity of PERK is necessary for growth factor localization in neurons, whether its main phosphorylation substrate, eIF2α, mediates this activity, and whether the PERK risk alleles associated with neurodegeneration alter the distribution of growth factors. In Aim 2 we will follow on genetic data indicating that PERK regulation of growth factor localization depends on components of the calcium homeostasis machinery, both membrane channels and cytosolic mediators. In both Aims we will combine genetic analyses of sensory neurons in worms and mice with biochemical analyses, because of the complementary insights provided by these approaches. This project will identify a novel cellular pathway that underlies the polarized secretion of growth factor by neurons and thus their functional connectivity. Understanding how PERK performs this novel protective function in neurons will open options of manipulating these pathways in presymptomatic AD and tauopathy models, and into delineating sensitizing vs protective mechanisms.