An amyloid-notch hypothesis for Alzheimer's disease.

For more than 20 years, the amyloid hypothesis has provided an important framework for Alzheimer's disease (AD) research, yet after 50,000 papers, the nonpathological function of beta-amyloid (Aβ) remains enigmatic. This mystery is compounded by an absence of gross abnormalities in amyloid precursor protein (APP)-deficient mice and zebrafish even though APP has been highly conserved throughout vertebrate evolution. Here, the author hypothesizes that vertebrate cells express APP and release Aβ as part of a mechanism to optimize blood vessel density with the metabolite removal needs of local tissue neighborhoods. High-gain feedback of Aβ production at the rate-limiting γ-secretase step reduces Aβ production and Notch activation. Notch inhibition causes endothelial cells to adopt a tip cell morphology that induces more highly branched blood vessels. In vivo, γ-secretase inhibitors block Notch signaling and induce dense capillary networks that are similar to those in the brains of AD patients and mice. Notch inhibition could also contribute to synapse loss by reducing EphB2 receptor expression. EphB receptors are critical for the maintenance of dendritic spine morphology, and deficiencies result in immature spines that lack synaptic activity. This revised amyloid-Notch hypothesis may also explain the disappointing results of recent clinical trials with γ-secretase inhibitors.