Tau seeds translocate across the cell membrane to initiate aggregation
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
Deposition of hyperphosphorylated misfolded tau is a hallmark of many neurodegenerative diseases, but the exact mechanisms by which misfolded tau spreads to adjacent areas of the brain are not known. In this manuscript, which will be of broad interest to cell biologists and neuroscientists, the authors suggest that tau fibrils that translocate directly through the cell membrane induce aggregation of cytosolic tau. While the results appear stunning, there are alternative explanations to the authors' hypothesis that require further investigation.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Cell Biology (eLife)
Abstract
Neurodegenerative tauopathies, including Alzheimer’s disease and related disorders, are caused by intracellular aggregation of tau protein in ordered assemblies. Experimental evidence suggests that tau assemblies propagate pathology across brain networks. Tau seeds enter cells through endocytosis but must access the cytoplasm to serve as templates for their own replication. The mechanism by which this occurs is unknown. To study tau uptake, we began with a whole-genome CRISPR knockout screen, which indicated a requirement vacuolar H + ATPase (v-ATPase) components. Treatment with Bafilomycin A1, an inhibitor of the v-ATPase, also reduced tau entry. We next tested direct modifiers of endolysosomal trafficking. Dominant-negative Rab5a expression uniquely decreased tau uptake, as did temporary cold temperature during tau exposure, consistent with a primary role of endocytosis in tau uptake. However, despite reducing tau uptake, these interventions all paradoxically increased intracellular seeding. Consequently, we generated giant plasma membrane vesicles (GPMVs), which cannot undergo endocytosis, and observed that tau fibrils and monomer translocated into the vesicles, in addition to TAT peptide, whereas transferrin and albumin did not. In every case, tau required binding to heparan sulfate proteoglycans (HSPGs) for cell uptake, seeding, or GPMV entry. These findings are most consistent with direct translocation of tau seeds across the lipid bilayer, a novel mechanism of entry into the cytoplasm.
Article activity feed
-
Evaluation Summary:
Deposition of hyperphosphorylated misfolded tau is a hallmark of many neurodegenerative diseases, but the exact mechanisms by which misfolded tau spreads to adjacent areas of the brain are not known. In this manuscript, which will be of broad interest to cell biologists and neuroscientists, the authors suggest that tau fibrils that translocate directly through the cell membrane induce aggregation of cytosolic tau. While the results appear stunning, there are alternative explanations to the authors' hypothesis that require further investigation.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
In this manuscript, Dodd et al. study the uptake and cytosolic release of tau fibrils. Based on a Crispr knock-out screen which identified that vacuolar H+ ATPase factors were required for tau fibril uptake, the authors focused on the role of endocytosis and endocytic acidification. Using genetic and pharmacological approaches, the authors made the surprising observation that endocytosis inhibition reduced tau fibril uptake but strongly increased fibril-induced tau seeding in the cells. Authors demonstrate that giant plasma membrane vesicles unable to undergo endocytosis still take up tau fibrils. Dodd and colleagues conclude that tau fibrils likely directly penetrate the lipid bilayer, which represents a novel entry pathway involved in tau seeding
The authors present some perplexing divergent effects of …
Reviewer #1 (Public Review):
In this manuscript, Dodd et al. study the uptake and cytosolic release of tau fibrils. Based on a Crispr knock-out screen which identified that vacuolar H+ ATPase factors were required for tau fibril uptake, the authors focused on the role of endocytosis and endocytic acidification. Using genetic and pharmacological approaches, the authors made the surprising observation that endocytosis inhibition reduced tau fibril uptake but strongly increased fibril-induced tau seeding in the cells. Authors demonstrate that giant plasma membrane vesicles unable to undergo endocytosis still take up tau fibrils. Dodd and colleagues conclude that tau fibrils likely directly penetrate the lipid bilayer, which represents a novel entry pathway involved in tau seeding
The authors present some perplexing divergent effects of genetic or pharmacological inhibition of the endo-lysosomal system on tau fibril uptake and seeding. Surprisingly, inhibition of endo-lysosomal acidification inhibited fibril uptake but increased seeding in biosensor cells. Likewise, inhibition of Rab5a, a GTPase required for early endosome trafficking strongly impaired fibril uptake, but highly elevated tau seeding in biosensors. Low temperature during fibril incubation inhibited fibril uptake but increased seeding when cells were subsequently shifted to a normal growth temperature. Authors conclude that these results most likely show that endocytosis is not required for seeing tau. Instead, tau fibrils might directly translocate through the membrane to induce seeding.
Clearly, and rightfully stressed by the authors, the experiments demonstrate that the amount of internalized tau fibrils does not correlate with the tau seeding, arguing that minute (and potentially undetectable) amounts of tau seeds can very efficiently seed tau aggregation in biosensor cells. The massive amounts of labelled fibrils taken up by basically 100 % of cells might obscure the actual relevant tau seeds taken up by alternative routes that result in cytosolic tau seeding. This could apply to tau fibrils translocating through the membrane, as suggested by the authors, this could also apply to minute amounts of fibrils that still manage to enter the endo-lysosomal pathway even when this pathway is compromised pharmacologically or genetically. Any impairment of the endo-lysosomal system, however, likely also reduces the lysosomal degradative capacity. As a consequence, tau seeds in the endo-lysosomal system will have a higher propensity to escape to the cytosol to induce seeding. While the findings presented here are indeed very interesting and the Crispr screen is very elegant and informative, experiments are not sufficient to explain the cellular mechanisms involved.
-
Reviewer #2 (Public Review):
This manuscript describes cell culture experiments to begin to understand this process. A CRISPR-Cas9 screen identified various genes which influence cellular uptake of Alexa-fluor labeled tau protein including components of the vacuolar H+ ATPase (v-ATPase) which is required for acidification of lysosomes, confirmed using genetic and pharmacologic approaches. Perhaps most interestingly, changes in the cellular uptake of tau upon manipulating the endolysosomal system were inversely correlated with downstream tau aggregation as measured using overexpression of FRET-based biosensors. Moreover, tau appeared to be able to penetrate directly across the plasma membrane as shown through analysis of giant plasma membrane vesicles (GPMVs), and this appears to be related to heparan sulfate proteoglycans. Overall, the …
Reviewer #2 (Public Review):
This manuscript describes cell culture experiments to begin to understand this process. A CRISPR-Cas9 screen identified various genes which influence cellular uptake of Alexa-fluor labeled tau protein including components of the vacuolar H+ ATPase (v-ATPase) which is required for acidification of lysosomes, confirmed using genetic and pharmacologic approaches. Perhaps most interestingly, changes in the cellular uptake of tau upon manipulating the endolysosomal system were inversely correlated with downstream tau aggregation as measured using overexpression of FRET-based biosensors. Moreover, tau appeared to be able to penetrate directly across the plasma membrane as shown through analysis of giant plasma membrane vesicles (GPMVs), and this appears to be related to heparan sulfate proteoglycans. Overall, the findings in this manuscript suggest that while tau can be internalized via the endolysosomal system, tau protein that is internalized through this route may not be the primary species that leads to downstream tau inclusion formation. This is supported by multiple complementary approaches (CRISPR screen, targeted knockout, pharmacologic inhibition, dominant negative studies). Rather, the authors suggest that direct translocation of tau aggregates/seeds across the plasma membrane, facilitated by binding to heparin sulfate proteoglycans, may be responsible for downstream tau aggregation/seeding.
One weakness is that most of the studies use non-neuronal cells, although a few findings were replicated in neuronal cultures. In addition, the main outcome measures are fluorescence-based assays dependent on tagged tau proteins and/or overexpressed tau fragments fused to fluorescent proteins. From a mechanistic standpoint, it is unclear why lysosomal deacidification would affect cellular endosomal uptake and decrease overall intracellular tau levels (as opposed to leading to an increase in endocytosed tau due to decreased lysosomal degradation). Also, while the manuscript provides good evidence suggesting that endolysosomal tau does not lead to downstream tau aggregation/seeding, the manuscript does not appear to directly assess whether translocation across the plasma membrane, as opposed to other mechanisms that have been proposed in the literature, is the primary mechanism that leads to downstream tau aggregation/seeding.
-
Reviewer #3 (Public Review):
In this manuscript, Dodd et al. measure the internalization of exogenous fluorescently-labelled tau by cultured HEK cells and iPSC-derived neurons, as well as the aggregation of fluorescent fusion proteins of the repeat domain of tau with the P301S mutation (tau RD) expressed in these cells. They find that inhibition or reduction of V-ATPases and Rab5A reduces tau internalization and increases tau RD aggregation, as does culturing the cells at cold temperatures. The authors also find that exogenous fluorescently-labelled tau is internalized by HEK cell-derived GPMVs. All conditions are dependent on HSPGs, which presumably act as cell-surface attachment factors, similar to their role in the attachment of viruses to the cell surface. Based on the involvement of V-ATPases and Rab5A in endocytosis, the authors …
Reviewer #3 (Public Review):
In this manuscript, Dodd et al. measure the internalization of exogenous fluorescently-labelled tau by cultured HEK cells and iPSC-derived neurons, as well as the aggregation of fluorescent fusion proteins of the repeat domain of tau with the P301S mutation (tau RD) expressed in these cells. They find that inhibition or reduction of V-ATPases and Rab5A reduces tau internalization and increases tau RD aggregation, as does culturing the cells at cold temperatures. The authors also find that exogenous fluorescently-labelled tau is internalized by HEK cell-derived GPMVs. All conditions are dependent on HSPGs, which presumably act as cell-surface attachment factors, similar to their role in the attachment of viruses to the cell surface. Based on the involvement of V-ATPases and Rab5A in endocytosis, the authors conclude that endocytosis of tau does not contribute to the aggregation of expressed tau. In addition, based on the lack of endocytosis in GPMVs, the authors conclude that tau can translocate across membranes and that this contributes to the aggregation of expressed tau.
The observation that conditions that decrease the overall internalization of exogenous tau can increase the aggregation of expressed tau suggests that multiple internalization routes exist, some of which are non-productive for the aggregation of expressed tau. This has important consequences for therapeutic strategies aiming to limit the internalization of tau. However, the conclusions that tau can translocate across membranes and that this contributes to the aggregation of expressed tau, whereas endocytosis of tau is non-productive for the aggregation of expressed tau, are not fully supported by the data.
Major comments:
1. There appear to be several alternative interpretations other than a reduction of endocytosis for the effects of perturbing V-ATPase and Rab5A function and culturing cells at cold temperatures. First, internalized tau was measured 4 h after the addition of exogenous tau to the cells. This seems like a long time for the study of endocytosis, which occurs in minutes. By 4 h, degradation of tau may have an effect on the amount of measurable internalized tau. This is important because, in addition to their roles in endocytosis, V-ATPases and Rab5A also have roles in protein degradation via the endolysosomal system. Similarly, culturing cells at cold temperatures for 4 h is expected to have many effects beyond the inhibition of endocytosis. In addition, the authors do not control for humidity and CO2 concentration, which could also affect their measurements. Perturbation of V-ATPases and Rab5A could also be exerting their effects by reducing the translocation of tau across endolysosomal membranes, instead of endocytosis. The authors found that the expression of dominant-negative dynamin increased the amount of internalized tau. Is this unexpected, given that dynamin is required for most forms of endocytosis and has been previously reported to be required for tau endocytosis (Wu et al. 2013. J. Biol. Chem. 288, 1856-1870; Falcon et al. 2018. J. Biol. Chem. 293, 2438-2451; Evans et al. 2018. Cell Rep. 22, 3612¬-3624)?2. It is difficult to draw parallels between the experiments using cells and those using GPMVs. The authors use 25 nM tau for cell experiments, but 500 uM tau for GPMV experiments. This is a huge difference in concentration. The authors should carry out the GPMV experiments using the same concentration of tau as in the cell experiments. 500 uM is also a very high concentration and raises the question of if the GMPVs are completely sealed. GMPVs have recently been reported to be permeable to hydrophilic macromolecules (Skinkle et al. 2020. Biophys. J. 118, 1292-1300). Tau and the TAT peptide are more hydrophilic than the two negative controls used, transferrin and albumin.
3. It is not clear which molecular species of tau (monomers, oligomers, or fibrils) are being studied. The authors refer to tau fibrils, but the species of recombinant tau they are using are never characterised. Incubation of tau with heparin can be expected to result in a mixture of fibrils, oligomers, and monomers. Sonication may also change the distribution of tau species by liberating oligomers and monomers from fibrils. Similarly, key details about the immunoprecipitation are lacking, including neuropathological characterization of the human cases, the brain region, the amount of brain tissue, the lysis buffer, the epitope of the Tau B antibody, the amount of Dynabeads, and analysis of the immunoprecipitated sample to show what species of tau are present.
-
