Distinct forms of amyloid-β moderate sleep duration through NAD+-linked redox metabolism in Alzheimer's disease.
Sleep disruptions precede a clinical diagnosis of Alzheimers disease (AD) by several years. However, how AD pathologies affect sleep remains unclear. Here, we integrate epidemiological data with insights from Drosophila models of AD to investigate how AD progression could be linked to sleep disruption. We found that individuals with a high risk of AD report a shorter sleep duration than do those with a clinical diagnosis of AD. We showed that the expression of different forms of amyloid-{beta} in flies can replicate these variations in sleep duration. Analysis of the metabolome and proteome of these flies revealed distinct changes in NAD+-linked redox metabolism and levels of hyperkinetic (Hk), a redox-sensing sleep homeostat. We showed that the genetic upregulation of Hk is neuroprotective and increased KCNAB2 expression in the brain decreases AD risk in humans. Overall, our data provides a new mechanism linking the disruption of NAD+-linked redox sensing and sleep disruption in AD.