ARF6, a new regulator of retromer function and cholesterol homeostasis in neurons.

点击翻译按钮获取中文摘要

SUMMARY We propose to extend our published work on the small GTP-ase Arf6 (ADP-ribosylation factor 6) and its control of retromer function and cholesterol homeostasis. Retromer dysfunction and cholesterol dyshomeostasis, although both suggested to be some of the earliest events in neurodegeneration, including in Alzheimer's disease (AD), have not been mechanistically linked in neurons. We recently provided evidence that, in fibroblasts, Arf6 ablation resulted in a selective increase of an endosomal pool of phosphatidylinositol-4- phosphate (PI4P). This caused a perturbation of retromer trafficking that led to mistrafficking of its cargoes, including cation-independent mannose-6-phosphate receptor (CI-M6PR) and its ligand NPC2, and to an accumulation of cholesterol in the late endosomes/lysosomes (LE/Lys). Here, we postulate that, in neurons, retromer function and cholesterol homeostasis actually belong to a single pathway regulated by Arf6. We will test our hypothesis using primary cultures of hippocampal neurons derived from wild-type (WT) mice or from the Arf6 flox/flox (fl/fl) mouse model developed in the lab. We will first test whether retromer function controls cholesterol homeostasis (Aim1) using WT hippocampal neurons infected with a control shRNA or an shRNA directed against Vps35, the core subunit of the retromer complex. We will examine whether cholesterol distribution is modified in Vps35-depleted neurons using filipin, a fluorescent ligand of cholesterol. To further provide mechanistic insights, we will test whether trafficking of the retromer cargo CI-M6PR and its ligand NPC2 are altered in Vps35-depleted neurons, using immunocytochemistry and confocal microscopy. We will further test whether Arf6 controls retromer function and cholesterol homeostasis (Aim2) using hippocampal neurons derived from Arf6 fl/fl mice and infected with a control or a Cre lentivirus. We will assess whether retromer function is impaired in the absence of Arf6 by analyzing the subcellular localization and dynamics of the retromer subunits as well as the fate of its cargoes using confocal and spinning disk microscopies. We will also examine cholesterol distribution in control and Arf6-depleted neurons using filipin. Finally, we will identify the effectors of Arf6 on retromer function (Aim 3) in hippocampal neurons. We will test whether Arf6 ablation affects the endosomal levels of PI4P in Arf6-/- neurons. In addition, we will characterize the enzymatic effector of Arf6 on endosomal PI4P levels using a shRNA screen of the different isoforms of phosphatidylinositol-4- phosphate 5-kinases, which phosphorylate PI4P into PI(4,5)P2, in WT neurons. We will then reintroduce the identified wild-type Arf6 effector -or its kinase dead mutant- in Arf6-/- neurons and test if it rescues PI4P endosomal accumulation, retromer dysfunction and cholesterol accumulation. If successful, results from this proposal will provide mechanistic evidence that, in neurons, retromer dysfunction and cholesterol dyshomeostasis, both linked to neurodegeneration, are part of a single pathway regulated by Arf6 and will thus warrant the further study of Arf6 in neurodegenerative disorders, especially in LOAD.