Molecular and cellular mechanisms underpinning disruptions to neuronal signalling and vulnerability in Alzheimer’s disease neurons and novel applications of machine learning (#4)
Disruption to neuronal signaling and alterations in neuronal excitability have been identified as early phenotypes in Alzheimer’s disease (AD). Induced pluripotent stem cell (iPSC) derived neurons generated from AD donors were used as an in vitro model to interrogate the mechanisms underlying dysfunctional signaling and neuronal activity and increased vulnerability. Using a combination of calcium imaging, patch clamp and microelectrode arrays, with molecular and cellular techniques and machine learning approaches, we identified alterations in neuronal signaling in AD neurons in neuronal cultures and organoids.
AD neurons exhibited an increased susceptibility to nitrosative/oxidative stress damage, with an altered distribution of the lipid soluble antioxidant alpha-tocopherol and decreased levels of the antioxidant enzyme, glutathione peroxidase. This decreased resilience to redox dyshomeostasis increased neuronal vulnerability to the cell death pathway, ferroptosis. Induction of ferroptosis significantly reduced cell viability, multiple phosphatidylethanolamine polyunsaturated fatty acid lipid species and abolished glutamatergic calcium signaling. Supplementation with alpha-tocopherol, which can prevent ferroptosis, reduced the proportion of spontaneously signaling AD neurons and partially protected against the loss of select fatty acid species, implicating alterations to the lipid membrane in the breakdown of neuronal calcium signaling. Following on from this, novel machine learning tools were applied to identify alterations in AD neuron activity and function that can be ameliorated by targeting specific cell signaling pathways.
Together, the data indicate that the AD neuronal signaling phenotype is linked to both neuronal function and survival.