Neurological diseases are a growing public health concern with no cures and few treatments. Inflammation in the central nervous system (neuroinflammation) is observed in most neurological diseases and is thought to contribute to neuropathology. Our long-term goal is to identify targets for selective regulation of pathological inflammation in the CNS. Endoplasmic reticulum (ER) stress in neuronal and glial cells is associated with most neurodegenerative diseases. ER stress also drives inflammation, however the underlying mechanisms and impact on disease are unknown. Our central hypothesis is that PERK signaling in astrocytes contributes to neuroinflammation. This will be tested in three specific aims. 1) Determine the molecular mechanisms of PERK-dependent signaling leading to neuroinflammatory gene expression. Using traditional cellular and molecular biology and hypothesis-driven transcriptomics, the mechanisms of PERK-dependent neuroinflammation will be elucidated. 2) Determine the mechanisms by which PERK deletion affects neuroinflammation in vivo. Using in vivo cell-specific RNA labeling and conditional knockout mice we will define the spatial and temporal occurrence of ER stress and inflammatory gene expression in astrocytes, and how astrocyte-specific knockout of PERK affects the neuroinflammatory disease of experimental autoimmune encephalomyelitis. 3) Determine the non-cell autonomous mechanisms of ER stress signaling in astrocytes. Using in vitro primary cell co-cultures and in vivo conditional knockout mice, the PERK-dependent effects of astrocytes on the function and viability of resident and infiltrating cells of the CNS will be established. This project is significant because it will define basic mechanisms driving neuroinflammation and establish a potential therapeutic target for immune modulation in the CNS.