Abstract
Tau pathology is a major hallmark of Alzheimer’s disease (AD) and related diseases, called tauopathies. While Tau is normally enriched in axons, somatodendritic missorting of the microtubule-associated protein is a key event in early disease development. Tau missorting promotes synaptic loss and neuronal dysfunction but the mechanisms underlying both normal axonal sorting and pathological missorting remain unclear. Interestingly, the disease-associated Tau brain isoforms show different axodendritic distribution, but the distinct role of these isoforms in health and disease largely unknown. Here, we aimed to identify domains or motifs of Tau and cellular binding partners that are required for efficient axonal Tau sorting, and we studied the differences of the isoform-specific Tau interactome. By using human MAPT-KO induced pluripotent stem cell (iPSC)-derived glutamatergic neurons, we analyzed the sorting behavior of more than 20 truncated or phosphorylation-mutant Tau constructs, and we used TurboID-based proximity labelling and proteomics to identify sorting- and isoform-specific Tau interactors. We found that efficient axonal Tau sorting was independent of the N-terminal tail, the C-terminal repeat domains, and the general microtubule affinity of Tau. In contrast, the presence of the proline-rich region 2 (PRR2) was necessary for successful sorting. Our interactome data revealed peroxisomal accumulation of the Tau N-terminal half, while axonal Tau interacted with the PP2A activator HSP110. When we compared the interactome of 0N3R- and 0N4R-Tau, we observed specific interactions of 0N4R-Tau with regulators of presynaptic exocytosis and postsynaptic plasticity, which are partially associated with AD pathogenesis, such as members of the CDC42 pathway and the RAB11 proteins, while 0N3R-Tau bound to MAP4 and other cytoskeletal elements. In sum, our study postulates that axonal Tau sorting relies on the PRR2 domain but not on microtubule affinity, and unravels a potential isoform-specific role in synaptic function and AD-related dysfunction.
Competing Interest Statement
The authors have declared no competing interest.