SYNAPTIC MEMBRANE CHOLESTEROL: CULPRIT OR ACCOMPLICE OF ALZHEIMER'S DISEASE?

Amyloid precursor protein (APP) is tied to Alzheimer's disease (AD). Its proteolytic products, β-amyloid peptides (Aβs), are the major constituents of amyloid plaque, the hallmark of AD. APP mutations have been found in familial AD (fAD), and mice carrying those mutations exhibit different degrees of AD-like pathologies. Most studies about fAD mutations had looked for changes in APP cleavage, Aβ production and aggregation, and led to the popular amyloid hypothesis. However, more and more evidence has suggested that Aβ aggregation is not the causal factor, calling for a better appraisal of APP's intrinsic trafficking and processing in neurons. Various functionalities proposed for APP often require it to be in specific membrane sections like the plasma membrane or intracellular membrane compartments. Earlier studies mostly used non-neuronal cells, so we decided to use primary neurons in order to understand APP's behavior in the context of neurons. More importantly, we constructed triple-fluorescent reporter for both APP and synaptic vesicles (SVs) to address APP's surface turnover and cleavage along with synaptic vesicles in live neurons. We systematically validated the fidelity of the APP and synaptic vesicle reporters. Moreover, it also reported the N- and C-terminal secretase cleavage of APP. Despite through investigation, we failed to detect any activity-associated changes in APP expression, trafficking, transportation and cleavage. However, we did recorded a delayed surface turnover of APP during synaptic vesicle recycling. Surprisingly, we discovered an intriguing link between membrane cholesterol and surface APP. We further tested if and how different APP mutations alters neuronal membrane cholesterol, APP trafficking and processing. We discovered an inverse relationship between presynaptic surface APP fraction and membrane cholesterol, which involves a direct binding between APP and cholesterol. Disruption of APP-cholesterol interaction caused synaptic swelling and neuronal loss in a dominant-negative manner. Furthermore, we tested a battery of fAD mutations and found a significant inverse correlation between membrane cholesterol and surface APP but not cleavage.