Search

Background: The intersection of genetic diversity, memory, and synaptic function is a critical component in better understanding age‐related cognitive decline. Diversity Outbred mice offer the opportunity to investigate cognitive aging across a genetically diverse population in a controlled lab environment. DO mice exhibit an appreciable amount of variance in Contextual Fear Memory and Acquisition (CFM, CFA), which can be linked to individual differences in genetic background (Ouellette A. et al, 2022, Cell Reports). Here, we investigate the role of dendritic spines as a mediator of individual age‐related changes in memory. Methods: 524 DO mice (289 female, 235 male) were aged to either 8 or 18‐months. All mice received contextual fear conditioning (CFC) at 8mo, while another suet received an additional recall test at 17mo and renewal at 18mo. After CFC a subset of brains from each cohort were collected for spine morphology analyses (females, n = 42). One CA1 pyramidal neuron from each brain was imaged and quantified for the density of thin, stubby, mushroom and/or filipodia classified spines in addition to spine volume. Results: While was not a decline in CFM or CFA between 8 and 18mo, we observed expectedly wide range in individual memory outcomes. Thin spine density significantly decreased between 8mo and 18mo, and stubby spine density increased. Spine volume across all spine types increased with age. Apical thin spine density explained 61% of the variance in CFM outcomes, while thin spine volume explained 79% and 43% of variance CFA in apical and basal dendrites respectively. Thin spine density, however, did not associate with CFA outcomes. Conclusions: We have linked dendritic spine density and morphology as a potential mediator of individual outcomes across a genetically diverse population. Our results suggest that there may be an age‐related conversion of thin to stubby spine types, and that mice with fewer thin spines are more likely to have better CFM outcomes, coinciding with reports of thin spines as dynamic "learning spines" rather than "memory storage" spines (Hayashi, Y. 2005, Neuron). We also show that the size of these "learning" thin spines may be more important memory acquisition than their total density. [ABSTRACT FROM AUTHOR]

Background: Cognitive decline in Alzheimer's disease (AD) correlates well with the extent of tau pathology and synapse or dendritic spine loss. Thus, studying synaptic processes may aid in the identification of new therapeutic targets that limit cognitive decline. The entorhinal cortex (EC) is an initial site of tau tangle formation, from which tau can propagate to other brain regions through synaptic connections. We developed a pipeline whereby dendritic spine density and morphology as well as synaptic tau measurements from the EC were incorporated into a synaptic proteomic network to identify mechanisms of synaptic pathology in AD patients. Method: Golgi‐stained dendrites from postmortem human BA28 EC were imaged using high‐resolution brightfield microscopy and digitally reconstructed in 3D for morphometric analysis. Synaptosome fractions were isolated from the same BA28 samples. Levels of synaptosome and insoluble phosphorylated tau were measured by ELISA. Additionally, liquid chromatography coupled with tandem mass spectrometry‐based proteomics was performed on synaptosomes. Weighted Gene Co‐Expression Network Analysis (WGCNA) was used to generate a synaptic network of protein co‐expression modules. Module expression was correlated with dendritic spine measurements and phospho‐tau levels to guide selection of proteins for validation in model systems. Validation experiments were performed in neurons and Neuro‐2a cells, using a CRISPR activation system for target gene upregulation. Result: Spine density in the EC is reduced in AD, but not in cognitively normal individuals with AD pathology (CAD) cases. AD cases exhibit more synaptosomal pS396 tau than CAD cases. Module expression was correlated with dendritic spine density and morphology. We identified TWF2 as a hub protein of Module 16, which is positively correlated with thin spine length. Twf2 upregulation increased thin spine length in cultured neurons. Module expression was then correlated with phospho‐tau levels. We identified PPP1R7 as a hub protein of Module 24, which is inversely correlated with synaptosomal pS396 tau in human EC. Conclusion: We demonstrate that incorporating orthogonal cellular and molecular measurements into a proteomic network enables unbiased identification of proteins functionally involved in regulating those traits. Beyond elucidating synaptic processes in AD, this pipeline offers a blueprint to better contextualize omics‐based targets with relevant biological mechanisms. [ABSTRACT FROM AUTHOR]