Conversely, treatment with protein translation inhibitors, cycloheximide or anisomycin, alone (Fig. S8A and B) almost completely abrogated MAPT missorting (schematic in Fig. 2E). These results suggest that the dendritic MAPT is locally generated.
Our previous study [23] had shown that trehalose induces autophagy in primary neurons and in an N2a cell model of tauopathy, and efficiently reduces the level of MAPT and MAPT aggregation.
Trehalose treatment indeed supressed the missorting of MAPT down to ~4±0.7% of dendrites, far less than the control level (~16% of dendrites) (Fig. 5B).
In order to determine the impact of protein degradation systems on the sorting of MAPT, we sought to suppress their activity by treating neurons on the neuritic side of the MFCs with either autophagy inhibitors, wortmannin [25] and bafilomycin A 1 [26], or with proteasomal inhibitors, epoxomicin and lactacystin [27,28].
Similarly, a dramatic elevation of ubiquitinated substrates (from 4.2±0.1 to 37±1) upon epoxomicin treatment was detected in dendrites containing MAPT (compare Fig. S4E and Fig. S4F).
Missorted dendritic MAPT showed phosphorylation mainly at the 12E8 sites upon treatment with either the autophagy inhibitor wortmannin (Fig. 4B; 57.2±9.4% dendrites) or the proteasomal inhibitor epoxomicin (Fig. 4C, 62.9±7.4% dendrites) (Fig. 4A-C, quantification in Fig. 4D), but not at the AT8 and the PHF1 (p-S396/p-S404) sites (Fig. S5, Fig 4D).
MAPT-free dendrites in controls had a spine density of ~17 per 20-µm (17.2±0.5 per 20 µm) length, and there was a ~2- fold decrease in spine density in the MAPT-containing dendrites in neurons treated with wortmannin (7.4±1.0 per 20 µm) or epoxomicin (8.2±1.7 per 20 µm) (Fig. 6D).
Treatment with rolipram (10 µM) on the neuritic side for 24 h (Fig. 5C) supressed missorting of MAPT down to 4.5±0.4% dendrites, far below the control level (20.5±5.6% dendrites) (Fig. 5D), thereby confirming the role of the proteasome in degrading dendritic MAPT.
Compared with the MAPT-free dendrites of control neurons (Fig. S4A, 1.8±0.1), a substantial increase of SQSTM1 level (21.1±0.9) was observed in MAPT-containing dendrites upon wortmannin treatment (Fig. S4B).
In addition, it has been reported that in cultured neurons, Aβ oligomers induce MAPT missorting into the somatodendritic compartment, and the missorted MAPT is phosphorylated mainly at the 12E8 (p-S262/p-S356) and AT8 (p-S202/p-T205) sites [6].
At more mature ages (DIV14 - DIV21), MAPT localization is mainly found in the axons with only basal levels in the somatodendritic compartment, consistent with earlier findings [20].
Our previous work showed that inhibition of either the proteasome or autophagy in primary neurons induces pronounced neurotoxicity [23,24], making it impossible to address this issue in conventional neuron cultures.
In controls, the fraction of MAPT-containing dendrites on the neuritic side was low (~14%), but local treatment with inhibitors of either autophagy (wortmannin [Fig. 3B], bafilomycin A 1 [Fig. S3A]) or the proteasome (epoxomicin [Fig. 3C], lactacystin [Fig. S3B]) caused a pronounced 4-to 5-fold increase of MAPT-containing dendrites (to levels of ~50-76%) (Fig. 3D).
Thus, the dendritic and axonal MAPT are differentially phosphorylated. Based on this observation, we can conclude that the dendritic MAPT degraded by autophagy or proteasomal pathways is phosphorylated mainly at the 12E8 site.
Previous studies had shown that the missorting of MAPT into dendrites can cause loss of spines [6,7].
MAPT in AD and other tauopathies is hyperphosphorylated [29] and the hyperphosphorylation has been proposed to drive the missorting of MAPT.
This observation, together with local synthesis of MAPT protein, confirms that Mapt mRNA is present in dendrites and is actively translated.
BEL Commons is developed and maintained in an academic capacity by Charles Tapley Hoyt and Daniel Domingo-Fernández at the Fraunhofer SCAI Department of Bioinformatics with support from the IMI project, AETIONOMY. It is built on top of PyBEL, an open source project. Please feel free to contact us here to give us feedback or report any issues. Also, see our Publishing Notes and Data Protection information.
If you find BEL Commons useful in your work, please consider citing: Hoyt, C. T., Domingo-Fernández, D., & Hofmann-Apitius, M. (2018). BEL Commons: an environment for exploration and analysis of networks encoded in Biological Expression Language. Database, 2018(3), 1–11.