Why Alzheimer’s is a Disease of Memory: Synaptic Targeting by Pathogenic Aβ Oligomers (ADDLs)

Early Alzheimer’s disease manifests as a crippling inability to form new memories, but why Alzheimer’s is specific formemory has yet to be answered. As evidenced by thismeeting, research increasingly focuses on deterioration of synapses and dendritic spines (1), a concept introduced more than 30 years ago by Scheibel and colleagues (2). This synaptic damage is now attributed to the impact of soluble Abeta oligomers, thanks to contributions from multiple laboratories. Abeta oligomers (here referred to as “ADDLs”) were identified in 1998 as a new type of toxin, structurally distinct fromamyloid fibrils, that rapidly prevented LTP (3). ADDL-induced disruption of plasticity ismeasurable atmultiple levels, including ectopic over-expression of Arc, a spine cytoskeletal protein essential for memory formation. Confirming predictions based on the Arc response, ADDLs cause critical receptors to be eliminated fromsynaptic membranes and induce aberrations in spine morphology, with sustained presence of ADDLs resulting in spine elimination (4). Themechanism underlying these synaptic pathologies likely holds the key to understanding whyADis specific formemory.ADDL-induced pathologies are not broad consequences ofwholescale neuronal deterioration but instead derive from a highly specific attachment to the spines of certain excitatory synapses.Whether obtained fromADbrain or prepared in vitro,ADDLsbind to their targeted spines with high affinity, essentially acting as gain-of-function pathogenic ligands. Brain-derived and synthetic ligands are structurally equivalent 12mers that are strikingly elevated in AD brain (5) and also appear in animal models of AD, roughly concomitant with memory failure. Recent investigations into the synaptic targets of ADDLs implicate NMDA receptors, insulin receptors, and neighboring synaptic proteins.Memantine, an NMDA receptor antagonist used as an AD therapeutic drug, effectively inhibits ADDL-induced pathologies in the short term, while antibodies against the NR1 subunit significantly reduce ADDL binding. These results suggest that ADDLs bind at or near NMDA receptors. Insulin receptors are implicated by findings that prior exposure of neurons to insulin results in virtually complete inhibition of ADDL binding. While unoccupied insulin receptors are essential forADDLbinding, they arenot sufficient, implying that high affinity binding depends upon additional co-receptors, possibly comprising NR1 subunits or other nearby proteins. A synaptic response that may prove especially important for cognitive failure is a rapid and massive removal of insulin receptors from dendritic plasma membranes triggered when ADDLs are added prior to exogenous insulin. Accompanying this removal is an increase of insulin receptors within neuronal cell bodies, a net receptor redistribution that renders neurons insulin resistant. Antagonistic interaction between ADDLs and insulin provides a basis for possible CNS insulin resistance in AD and predicts therapeutic benefits for drugs that promote brain insulin function. Overall, knowing how ADDLs target and disrupt specific synapses will bring us closer to understanding why AD is a disease of memory and provide new avenues for the discovery of disease-modifying therapeutic drugs.