Alterations in Synaptic Structure and Tau Phosphorylation in Alzheimer's Disease

Abstract

Alzheimer’s disease (AD) is the most common cause of dementia in individuals over the age of 65. Of the two hallmark pathologies of AD, tau neurofibrillary tangles (NFTs) correlate more strongly with cognitive decline than do Aβ plaques. NFTs form in the entorhinal cortex (EC) and appear in anatomically connected regions over time as disease progresses. This tau spreading may occur via synaptic connections. Most excitatory synapses occur on dendritic spines, which house the postsynaptic machinery. Spines are actin-rich, plastic structures that change their size and shape in response to synaptic activity in a process that is dependent on actin regulation. In AD, spine density is reduced in several brain regions. However, individuals who are cognitively normal despite harboring AD pathology do not have reductions in dendritic spine density in dorsolateral prefrontal cortex (DLPFC). This suggests that maintenance of dendritic spines or synapses confers resilience to AD, in line with work demonstrating that synaptic density correlates with cognitive function more strongly than either tau NFTs or Aβ plaques do. Thus, synapses are important in AD as they may be the route of tau translocation throughout the brain, and maintenance of spines correlates strongly with cognitive function. Here, we explored the relationship between synaptic phospho-tau and dendritic spine density and morphology in a tauopathy mouse model. We did not observe a direct role of synaptic phospho-tau in spine loss, suggesting that synaptic accumulation of phospho-tau may precede dendritic spine loss or may affect synaptic function rather than structure. To determine whether dendritic spines are mediators of cognitive resilience in the EC, we assessed dendritic spine density and morphology in postmortem human EC. As in DLPFC, spine density was not reduced in cognitively resilient individuals, indicating that EC spine maintenance confers resilience to AD pathology. Additionally, we used an integrative protein co-expression network to identify regulators of dendritic spines. Using this approach, we identified twinfilin-2 (TWF2) as a potential modulator of thin spine length, demonstrating for the first time a role of TWF2 in regulating dendritic spine actin dynamics. This work furthers our understanding of synaptic structure and tau phosphorylation in AD.