Neuronal Selection Based on Relative Fitness Comparison Detects and Eliminates Amyloid-β-Induced Hyperactive Neurons in Drosophila

During adult life, damaged but viable neurons can accumulate in the organism, creating increasingly heterogeneous and dysfunctional neural circuits. One intriguing example is the aberrant increased activity of cerebral networks detected in vulnerable brain regions during preclinical stages of Alzheimer's disease. The pathophysiological contribution of these early functional alterations to the progression of Alzheimer's disease is uncertain. We found that a unique cell selection mechanism based on relative fitness comparison between neurons is able to target and remove aberrantly active neurons generated by heterologous human amyloid-β in Drosophila. Sustained neuronal activity is sufficient to compromise neuronal fitness and upregulate the expression of the low fitness indicators Flower(LoseB) and Azot in the fly. Conversely, forced silencing of neurons restores brain fitness and reduces amyloid-β-induced cell death. The manipulation of this cell selection process, which was already proved to be conserved in humans, might be a promising new avenue to treat Alzheimer's.