O3–02–01: Reversal of Synaptic Plasticity and Spatial Memory Deficits in APP/PS1 Mice by Suppressing Protein Kinase GCN2

Tao Ma,Eric Klann
DOI: https://doi.org/10.1016/j.jalz.2013.04.242
2013-01-01
Abstract:The beta-amyloid hypothesis of Alzheimer's disease (AD) has resulted in current research focusing on lowering levels of brain beta-amyloid, a small peptide derived from amyloid precursor protein. Meanwhile, the signaling pathways downstream of beta-amyloid as well as beta-amyloid-independent mechanisms are also being vigorously pursued as potential disease-modifying targets. One such potential target is the protein kinase GCN2, which is important in controlling mRNA translation via phosphorylation of eukaryotic initiation factor 2α subunit (eIF2α). Interestingly, elevated eIF2α phosphorylation has been correlated with AD pathogenesis, potentially via mechanisms of oxidative stress. Here taking advantage of a mutant mouse model in which eIF2 a kinase GCN2 is deleted (GCN2 KO), we examined whether repression of eIF2α phosphorylation could alleviate synaptic and memory deficits associated with APP/PS1 AD model mice. Hippocampal long-term potentiation (LTP) was induced by high frequency-stimulation (HFS) at CA3-CA1 synapses. Spatial memory was estimated by doing Morris water maze. Immunofluorescence combined with confocal microscopy was used to detect dendritic co-localization. Western blot was used to examine regulation of protein levels. We found that phosphorylation of eIF2α was increased in human AD brains and in the hippocampus of APP/PS1 mutant mice, which are a model for AD. We also found that the increase in eIF2 a phosphorylation in the APP/PS1 mice was correlated with increases in the levels of GCN2 and ATF4 (activating transcription factor 4). Additionally, we found that GCN2 co-localized with phospho-eIF2α in dendritic processes of hippocampal neurons. In correlation, hippocampal LTP was intact in GCN2 KO mice treated with beta-amyloid 1–42 (500 nM), which caused LTP failure in wild type mice. Furthermore, we crossed APP/PS1 mouse with GCN2 KO mice and generated APP/PS1/GCN2 KO double mutant mice. We demonstrated that inhibition of eIF2α activity by deleting GCN2 prevented decline of spatial memory and synaptic function in APP/PS1 mouse (10–12 month old). Our findings indicate that AD-associated impairments in synaptic plasticity and memory can be prevented by decreasing GCN2-dependent eIF2 a phosphorylation, which might be linked to enhanced translation. These studies suggest that GCN2 and eIF2a are viable therapeutic targets for synaptic dysfunction in AD and other related dementia syndrome.
What problem does this paper attempt to address?