DESCRIPTION (provided by applicant): Alexander disease (AxD) patients carry heterozygous mutations within the coding region of GFAP. To facilitate mechanistic studies on the pathogenesis of AxD, and provide animal models suitable for testing potential therapies, our long-term goal has been to generate mouse models for this disorder. We have now generated knock-in lines of mice carrying several of the most common GFAP mu-tations found in human AxD (equivalent to R79H, R239H, R239C, and R416W), and found that expression of the mutant GFAPs induces Rosenthal fibers (the hallmark pathologic feature), but the mice are viable. Moreover, expressing the mutant GFAPs in the context of appropriate genetic modifiers (such as elevating wild type GFAP) results in a lethal phenotype. In addition, altering GFAP expression either by simple over-expression or production of mutant GFAPs leads to induction of multiple stress pathways (aB-crystallin, Nrf2) that suggest specific hypotheses about pathogenesis and, ultimately, strategies for therapy. With the goals of understanding more precisely the consequences of abnormal GFAP expression in astrocytes, and of generating mouse models that reflect key phenotypic features of AxD in humans, we propose the following specific aims: In Aim 1 we will test the consequences of expressing mutant GFAP protein and/or GFAP over-expression, for comparison with known features of the Alexander phenotype in humans. These studies will include both in vitro models using primary cultures of mouse astrocytes, as well as the in vivo knock-in models carrying point mutations in the endogenous mouse GFAP gene. A range of properties will be examined, including biochemical, functional, and behavioral. Reversibility will also be assessed. In Aims 2 and 3 we will test the roles of the small stress protein aB-crystallin, and the transcription factor Nrf2, respectively, as modifiers, by crossing the GFAP mice with knockouts or with newly generated transgenics that over-express the two genes. Specific mechanisms by which excess aB-crystallin or Nrf2 might achieve rescue will be evaluated. These studies promise novel information on the pathological significance of mutant intermediate filament expression in astrocytes, will suggest mechanisms by which primary astrocyte dysfunction leads to generalized CMS disease, and will identify critical stress pathways that could ultimately serve as the basis for therapeutic interventions to mitigate the devastating effects of this disease.