In our hands, these aggregated tau species formed in different conditions did not show any significant release of LDH when applied on the differentiated SH-SY5Y cells (Supplementary Fig. 8A), and they did not show a significant reduction in cell viability by the MTTassay (Supplementary Fig. 8B).
The fibril-treated cells (cross-linked with GA or not) did not reduce the spine density significantly (Fig 4C, bars 8 and 9).
Recently, extracellular tau oligomers were shown to impair long term potentiation (LTP) and memory [40].
For comparison, in the case of Ab oligomers, only higher concentrations (w1 mM) cause an appreciable spine reduction of w30% (Fig. 4C, bar 3) (Zempel et al., 2010 [30]).
We reasoned that microglia and other cell types in the hippocampus might act as mediators for the cytotoxicity caused by TauRDΔK oligomers, which could explain why cytotoxicity was not observed in the cell culture systems.
We indeed found that there is an overexpression (w15%) of Nox1 protein (a component of NADPH oxidase complex) by Western blot, suggesting the role of NADPH oxidase complex as a potential source of ROS
These findings suggest that the ROS production induced by extracellular TauRDΔK oligomers might cause the activation of NADPH oxidase complex
TauRDΔK comprises the structural elements required for the pathologic assembly of tau filaments, and it causes reversible memory deficits and synapse loss in regulatable transgenic mice [11,25].
Analysis of sarkosyl extracts of brain homogenates of mice expressing the pro-aggregant repeat domain TauRDΔK revealed that oligomers are present, partly in a disulfide–cross-linked form (Supplementary Fig. 1).
During early stages of assembly, there is some increase in ThS intensity, combined with a pronounced increase in ANSfluorescence (Fig. 2Aand B), which is due to oligomers, indicating a change in conformation without increase in beta-structure
As shown earlier, TauRDΔK-expressing mice display loss of neurons in the CA3 and other regions of the hippocampus [11,24].
By contrast, monomers of TauRDΔK even at 10 mM concentration did not cause any significant ROS increase (Fig. 5B).
By contrast, there was no significant increase in the calcium level in cells treated with TauRDΔK monomers even at higher concentration (10 mM) (Fig. 5E and F, green curves).
During the transition from oligomers to polymers, the ANS fluorescence remains roughly constant, whereas ThS fluorescence increases strongly (Fig. 2B)
To address this question, we first applied the oligomers directly after the purification of the protein eluting from the Butyl FF 16/ 10 column (without buffer exchange; 1, 5, and 10 mM) to SHSY5Y cells and observed a variety of toxic effects, including pronounced reduction in the cell viability (by MTT assay, Fig. 3A), increase in apoptotic cells (by Hoechst staining, Supplementary Fig. 4A), loss of mitochondrial membrane potential (by JC1 assay, Supplementary Fig. 4B), caspase 3/7 activation (Supplementary Fig. 4C-D), and cytochrome-c release (Supplementary Fig. 4), within 5 hours of incubation.
Consistent with this, NeuN staining of the slices fixed after 48 hours of treatment with TauRDΔK oligomers revealed no reduction in the neuronal number in all regions of the hippocampus (CA1, CA3, and DG) (Fig. 3C and D, Supplementary Fig. 7), confirming that TauRDΔK oligomers do not cause cell death in the OHSC model as well.
There was no significant change in the LDH release in oligomer-treated cells compared with controls indicating that TauRDΔK oligomers do not compromise the membrane integrity (Supplementary Fig. 5G-H).
In this case, there was a dramatic loss (up to 50%) of spines in TauRDΔK oligomer-treated cells compared with monomer-treated cells (Fig. 4A and B, bar 3).
Drebrin, a neuronal actin-binding protein involved in spinogenesis and synaptogenesis, was decreased by up to 60% consistent with the reduced number of spines (Fig. 4D, bars 3, 6, and 9).
Endogenous tau protein retained its normal axonal localization and did not missort into the cell body and dendrites (Fig. 6A8), although there was a reduction in the spine density (Fig. 6B6) in the TauRDΔK oligomer-treated neurons.
The results revealed that both TauRDΔK and TauFLΔK oligomers reduce the density of spines up to 50% compared with the buffer-treated cells.
PSD-95, a marker of postsynaptic spines, decreased up to 50% in the TauRDΔK oligomer-treated cells, compared with buffer- and monomer-treated cells
Similarly, the GluR1 subunits of a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors (characteristic of mature spines and necessary for LTP and calcium signaling) decreased in the oligomer-treated samples up to 60%.
We also looked at the presynaptic protein synaptophysin, which was not significantly altered in TauRDΔK oligomer-treated cells compared with buffer- or monomer-treated cells (Fig 4D, bar 12). However, at higher concentration, synaptophysin tends to be reduced.
We observed an oligomer-dependent increase in the ROS production in the mature rat primary hippocampal neurons in all cellular compartments (Fig. 5A).
However, we did not find significant changes in the expression level of Nox2 protein (another component of NADPH oxidase complex) (Fig. 5C).
The ratio of 340 to 380 nm in TauRDΔK oligomer-treated cells showed a steady concentration-dependent increase in the intracellular calcium with a maximum reached at 20 minutes of incubation with oligomers (10 mM) occurring in all cell compartments (Fig. 5D; arrows).
We did not observe any increase in the phosphorylation of the tau repeat domain (as seen by the antibody 12E8) after treating the neurons with TauRDΔK oligomers.
We also checked the phosphorylation of tau at other sites (e.g., using the antibody AT8, reacting only with the endogenous tau) and did not observe an increase in the phosphorylation.
Tau oligomers from TauRDΔK and TauFLΔK mice reduced the density of the synapses by w50%, whereas tau from wild-type mice had no effect on the density (Fig. 7G).
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.