Dual truncation of tau by caspase-2 accelerates its CHIP-mediated degradation.

Intraneuronal aggregates of the microtubule binding protein Tau are a hallmark of different neurodegenerative diseases including Alzheimer's disease (AD). In these aggregates, Tau is modified by posttranslational modifications such as phosphorylation as well as by proteolytic cleavage. Here we identify a novel Tau cleavage site at aspartate 65 (D65) that is specific for caspase-2. In addition, we show that the previously described cleavage site at D421 is also efficiently processed by caspase-2, and both sites are cleaved in human brain samples. Caspase-2-generated Tau fragments show increased aggregation potential in vitro, but do not accumulate in vivo after AAV-mediated overexpression in mouse hippocampus. Interestingly, we observe that steady-state protein levels of caspase-2 generated Tau fragments are low in our in vivo model despite strong RNA expression, suggesting efficient clearance. Consistent with this hypothesis, we find that caspase-2 cleavage significantly improves the recognition of Tau by the ubiquitin E3 ligase CHIP, leading to increased ubiquitination and faster degradation of Tau fragments. Taken together our data thus suggest that CHIP-induced ubiquitination is of particular importance for the clearance of caspase-2 generated Tau fragments in vitro and in vivo.

Memory trace interference impairs recall in a mouse model of Alzheimer's disease.

In Alzheimer's disease (AD), hippocampus-dependent memories underlie an extensive decline. The neuronal ensemble encoding a memory, termed engram, is partially recapitulated during memory recall. Artificial activation of an engram can restore memory in a mouse model of early AD, but its fate and the factors that render the engram nonfunctional are yet to be revealed. Here, we used repeated two-photon in vivo imaging to analyze fosGFP transgenic mice (which express enhanced GFP under the Fos promoter) performing a hippocampus-dependent memory task. We found that partial reactivation of the CA1 engram during recall is preserved under AD-like conditions. However, we identified a novelty-like ensemble that interfered with the engram and thus compromised recall. Mimicking a novelty-like ensemble in healthy mice was sufficient to affect memory recall. In turn, reducing the novelty-like signal rescued the recall impairment under AD-like conditions. These findings suggest a novel mechanistic process that contributes to the deterioration of memories in AD.