Nuclear tau aggregates inhibit RNA export and form by secondary seeding from cytosolic tau aggregates

  1. Carolyn J. Decker
  2. Kathleen McCann
  3. Evan Lester
  4. James Pratt
  5. Meaghan Van Alstyne
  6. Yuchen Wang
  7. Roy Parker

Reviewed by

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

Log in to save this article

Abstract

Tau aggregates contribute to multiple neurodegenerative diseases including frontotemporal dementia and Alzheimer’s disease (AD). In models of tauopathy and in patient tissue, tau aggregates can form in the cytoplasm, perinuclear region, and nucleus. Using a HEK293T tau biosensor system, we identified that cytoplasmic tau aggregates formed first, followed by perinuclear-ring-like tau assemblies, and then nuclear tau aggregates formed in nuclear speckles. Nuclear tau aggregates only form in cells with pre-existing cytoplasmic tau aggregates and mostly form independently of cells traversing mitosis. Finally, nuclear tau aggregates do not contain exogenous tau seeds and arise by a secondary seeding event dependent on VCP. Nuclear tau aggregates inhibit mRNA export and show a twofold increase in poly-adenylated mRNAs in the nucleus. Together, these findings indicate that nuclear tau aggregation alters RNA biogenesis and occurs by a secondary seeding event from cytoplasmic tau aggregates, which could contribute to tau pathology.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    Decker et al present an interesting study of the order of events in tau seeding in a biosensor HEK293 derived cell line. This a critical unresolved question in the field about subcellular compartment contributions to tau aggregation. This exploration of the nuclear tau aggregated deposition and seeding in HEK293T tau biosensor cells uses a variety of imaging-based methods. They show that nuclear aggregates only form in cells with cytosolic aggregates, nuclear aggregates cannot form in the absence of cytoplasmic tau aggregates. The original tau seeds do not persist. Also aggregates in the nucleus are dependent on VCP and SRRM2 for formation. The paper is limited in scope by use of only HEK239T cells and seem to overstate the generalizability of the findings to neuronal mechanisms of neurodegeneration. Please see to all tauopathies. In particular, the manuscript does not outline the overstatement of some of the conclusions.

    Key Points to address:

    1. The manuscript does not detail limitations of the study in the discussion. Please address the concern that HEK293T biosensor cells are not neurons. Especially in the clear animations showing the transformation from cytoplasmic to nuclear aggregates appears to require cell division and nuclear breakdown.

          We agree with the reviewer that a limitation of this manuscript is we only used HEK293 cells.  We have added text to emphasize this point in a "Limitations of this study" section at the end of the discussion.  However, as a starting point we believe understanding the cell biology of protein aggregation even in non-neuronal cells can be of value. 

    Moreover, we clearly see cases of nuclear tau aggregates forming without cell division and nuclear breakdown (Figure 1 and Movies).  We have added text to emphasize this point since it is relevant to the potential formation of nuclear aggregates in neurons and the reviewer must have missed this point.

    The introduction sets this up as Alzheimer's disease relevant but all studies are down with P301S tau which is a distinct and particularly aggressive form of tauopathy (FTLD-Tau). There is no amyloid beta component to any of these studies.

            This is a good point, and we have clarified our use of an FTLD model.  We do note that since seeds from post-mortem tissues in several different tauopathies can give nuclear tau aggregates (Sanders et al 2014), we anticipate that this process is general to multiple tauopathy contexts.

    The study does not address the peculiar structure of P301S aggregates, which while disease relevant are clearly distinct from AD or most forms of familial FTLD. The authors should limit the generalizability of the findings to their particular form of tauopathy unless they plan to use multiple tau fibril conformations in their studies.

            The reviewer points out that we have only used one model system, and presumably only one tau fibril structure and therefore we should be cautious about the generality of our results.  This is a valid point, and we now point out this limitation in the manuscript.

    The authors do not address the potential impact of fusing a natively unfolded protein like tau to a highly structured beta barrel like GFP. Please present this potential confound.

            We have added text pointing out that using GFP fusion proteins has the potential to alter tau function. We note this is an issue in the use of any fusion proteins, which have nevertheless proven useful tools.

    1. Inhibition of VCP can cause proteinopathies in the absence of other seeding. For instance, familial mutations in human VCP can cause either tau or TDP-43 proteinopathy depending on the specific human disease causing mutation. Thus, critical controls are missing from figure 3. For instance, the consequence of VCP inhibition on unseeded biosensor cells is a missing control. Second all panels should evaluate TDP-43 aggregation to ascertain whether or not the secondary nuclear seeding involves TDP-43.

          In this comment, the reviewer asks that we show the effects of VCP inhibition on unseeded cells.  We will add this control, and we observe no appreciable tau aggregation with tau seeding. 

    We will also assess whether TDP-43 aggregates in the HEK293 biosensor cells with or without VCP inhibition and/or tau seeding.  However, we note that it is clear from many studies that tau aggregation can occur independently of TDP-43 aggregation.

    Minor concerns: A. Line 635 - In line 380, they discuss that aggregation of tau does not lead to perturbations in nuclear transport. In line 390, they discuss that aggregation of tau does not affect nuclear envelope integrity or nuclear import. However, in the discussion discusses that aggregation alters nuclear RNA export. These statements could use clarifying that protein export is not perturbed but RNA export and import may be.

    We have clarified this point.

    B. Line 564: "This observation suggests that tau aggregation in the cytoplasm may lead to increased expression of some RNAs." This could also be that cytoplasmic tau alters RNA export. These experiments don't differentiate between these options.

    This comment is related to other comments about the relative abundance of specific RNAs in the nucleus or cytoplasm. We will add new data to the manuscript where we examine the numbers of specific RNAs in cells with and without nuclear or cytoplasmic tau aggregates. This will allow us to determine if there is simply a retention of RNAs in the nucleus or if, in some cases, there is also an increase in RNA levels.

    In Figure 1, the authors show large aggregates overlapping the nucleus. It is unclear whether these aggregates have a portion both within and outside the nucleus or if they are deforming the nucleus and are wholly external to the nuclear compartment. Clarity on this issue is important. If the nucleus is deferment the observed aggregates seem reminiscent of aggresome formation. Please clarify. We assume the reviewer asks us to clarify why the large cytoplasmic tau aggregates are localized near the nucleus. Indeed, we suspect these are accumulating in aggresomes over time and have added this point to the text. Importantly, we do not observe a general defect in the integrity of the nucleus suggesting that even those these assemblies are close to the nucleus, they are not altering the nuclear envelope. We have added text to explain this issue.

    Reviewer #1 (Significance (Required)):

    Decker et al present an interesting study of the order of events in tau seeding in a biosensor HEK293 derived cell line. This a critical unresolved question in the field about subcellular compartment contributions to tau aggregation. This exploration of the nuclear tau aggregated deposition and seeding in HEK293T tau biosensor cells uses a variety of imaging-based methods. They show that nuclear aggregates only form in cells with cytosolic aggregates, nuclear aggregates cannot form in the absence of cytoplasmic tau aggregates.

    Reviewer #2 (Evidence, reproducibility and clarity (Required)):

    The manuscript by Decker et al. examines the formation of nuclear tau aggregates and their functional consequences using a HEK293T tau biosensor system. The authors propose that nuclear tau aggregates arise through VCP dependent secondary seeding from cytoplasmic aggregates and that these nuclear aggregates impair RNA export. The study addresses an important and relatively unexplored aspect of tau biology. However, several conclusions extend beyond what the data directly supports, and several essential controls are missing. Major Comments

    • The introduction is generally clear and well organized. However, it would benefit from additional mechanistic context explaining how polyserine domains might promote tau aggregation and why this feature is biologically relevant.

           We have added text expanding what is known about how polyserine domains can increase tau fibrillization.

    • The live cell imaging convincingly demonstrates the temporal sequence of cytoplasmic followed by nuclear aggregation. However, the manuscript lacks controls assessing whether nuclear aggregation correlates with higher tau expression levels. Such controls are necessary to exclude expression driven artifacts.

           We will add an analysis of the relationship between tau expression levels and cells with nuclear tau aggregates.  We observed that tau aggregates were independent of the tau expression levels, ruling out that nuclear tau aggregates are solely an artifact of extremely high tau expression levels.

    • The authors conclude that nuclear envelope integrity is preserved, but only import assays were performed. To validate the sensitivity and specificity of the assay, export assays or positive controls for nuclear transport disruption are required.

           We had already shown that in cells with nuclear tau aggregates the nuclear export of mRNAs is perturbed. We will add additional analyses of whether nuclear export of proteins is altered.

    • The Cy3/Cy5 seed experiments support the claim that exogenous seeds do not enter the nucleus. However, the conclusion that VCP generates secondary seeds is overstated. For example, the manuscript states: "VCP is responsible for the formation of secondary seeds..." (lines 418-439), yet the data demonstrate correlation rather than direct evidence of seed generation.

           This is a valid point. We have rephrased the manuscript to note that VCP is required for nuclear tau aggregation, possibly through the formation of secondary tau seeds, which is consistent with earlier work suggesting VCP can generate new tau seeds (Saha et al., 2023, Nature Communications; Batra et al., 2025, Molecular Neurodegeneration).

    To substantiate this conclusion, the authors should: directly quantify seed abundance. The current interpretation assumes uniform cytoplasmic uptake of seeds but does not measure it; Include controls addressing VCP inhibitor specificity, as these compounds have pleiotropic effects (e.g., ER stress, proteostasis collapse). No data is provided on whether VCP inhibition alters tau ubiquitination, which could have major implications on tau aggregation.

                  This comment addresses the issue of whether VCP can generate new seeds from tau fibers.  This is a conclusion already reached by prior work (Saha et al., 2023, Nature Communications; Batra et al., 2025, Molecular Neurodegeneration).  The point of our manuscript that this comment addresses is whether the nuclear aggregates are forming from a secondary seeding event, for which we have already provided several lines of evidence.  First, we have shown that nuclear aggregates only form after the formation of a prior cytoplasmic tau aggregate (Figure 1). Second, we have shown that nuclear aggregates do not contain exogenous seeds, while all cytoplasmic tau aggregates do (Figure 2).  Finally, we have shown that nuclear tau aggregates are dependent on VCP, which is consistent with the prior work showing VCP can generate tau seeds.  It is beyond the scope of this manuscript to determine in more detail how VCP affect tau aggregates generally.  For this reason, and since we have robustly demonstrated our conclusion, we have chosen not to pursue these additional suggested experiments.

    • The authors observed increase in nuclear poly(A)+ RNA and specific transcripts. However, the current data do not distinguish between several possible mechanisms that may account for this increase, including impaired export, increased transcription, enhanced RNA stability, or nuclear retention due to speckle reorganization.

           To address this comment, we will quantify the levels of individual RNAs in the nucleus, cytoplasm and whole cell.  This will allow us to determine if there is an increase in RNA levels (possibly due to increased transcription or reduced decay), or if the increased nuclear RNA levels are due to block to mRNA export.  We will also assess transcription rate by measuring the intensity of the transcription loci, which will allow us to distinguish if any changes in mRNA levels are due to transcription or changes in RNA decay.

    • The discussion occasionally overinterprets the data. Several statements should be reframed as hypotheses rather than conclusions:

    1. "VCP can generate tau seeds capable of additional seeding within a cell." (lines 572-594) This has not been directly demonstrated and should be softened accordingly.

    We have done so.

    1. Active import via SRRM2 is proposed, but no supporting data are presented. This should be clearly framed as a speculative model.

    We have done so.

    1. "Tau aggregates in the nucleus alter the function of nuclear speckles..." (lines 616-637). While plausible, this is not directly shown. Alternative explanations such as transcriptional upregulation or stress induced changes should be acknowledged.

    We have altered this text to be more accurate.

    1. The statement "It is possible that such nuclear aggregates could alter nuclear RNA export and contribute to pathology." (lines 637-655) is reasonable, but the authors should emphasize that nuclear tau aggregates are not consistently observed across tauopathies and that the HEK293T biosensor system may not fully recapitulate neuronal biology.

    We agree with this point and have rephrased the text accordingly.

    Reviewer #2 (Significance (Required)):

    The manuscript by Decker et al. examines the formation of nuclear tau aggregates and their functional consequences using a HEK293T tau biosensor system. The authors propose that nuclear tau aggregates arise through VCP dependent secondary seeding from cytoplasmic aggregates and that these nuclear aggregates impair RNA export. The study addresses an important and relatively unexplored aspect of tau biology. However, several conclusions extend beyond what the data directly supports, and several essential controls are missing.

    Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    In their manuscript, Decker et al., investigate the relationship between cytoplasmic and nuclear tau aggregation using a HEK293T biosensor system and propose a multistep model in which cytoplasmic aggregates give rise to nuclear tau aggregates, potentially via VCP-dependent secondary seed generation and involvement of nuclear speckle components. The study further explores functional consequences of nuclear tau aggregation on RNA metabolism. Overall, the work is interesting and potentially impactful. The combination of live-cell imaging, seed-labeling strategies, genetic perturbations (SRRM2/PNN), and RNA imaging represents a thoughtful experimental approach. However, I have some minor concerns and feel the authors should address these -

    1. Poly(A)+ FISH intensity is not a direct measure of export efficiency. The authors claim that tau aggregation within nuclear speckles interferes with nuclear export of RNA. It is highly possible that increased nuclear RNA levels observed could reflect altered transcription, stability, or stress responses rather than export defects alone. In the case of ATF3, a known stress responsive gene, increased nuclear signal could reflect transcriptional activation, not export defects. To prove that export is defective, the authors should at least measure total RNA levels (qPCR) in nuclear vs cytoplasmic fraction.

          To address this issue, we will quantify the levels of specific RNAs in the nucleus and cytoplasm by smFISH, which will allow us to clarify why there are more RNAs associated with nuclear speckles in the context of nuclear tau aggregates.

    Though the authors have shown the proposed role of VCP in generating secondary seeds by using inhibitors, the authors should show genetic validation by using dominant-negative VCP.

            This experiment essentially asks us to examine the role of VCP in nuclear tau aggregation by an additional method. We will add experiments examining how nuclear tau aggregates form when VCP is knocked down by siRNAs.  We have chosen not to use dominant negative VCP mutants since their phenotype will be complicated with the endogenous VCP possibly remaining functional.

    **Referees cross-commenting** *This session contains comments from different reviweers* Reviewer 3 I agree with the reviewers that additional controls and experiments would strengthen the VCP inhibition studies. However, I would like to clarify that the specific concern raised by Reviewer 1 (Key point number 4) regarding fusion of tau to GFP does not apply to this manuscript. In this study, the authors use tau conjugated to Cy3, a well established approach in the field that adds only approximately 1 kDa to the protein.

    Reviewer 1 Apologies reviewer 3, but I respectfully disagree. Please look again at the legends for figs 1 through fig 5. All clearly delineate the use of tau biosensor cells using a YFP rather than GFP fusion protein with tau. i do agree we should correct my review to state YFP rather than GFP, but structurally the concern remains the same. Cy3 labelling, I believe is used to track the relatively short lived exogenous seeds.

    Reviewer #3 (Significance (Required)):

    The integration of approaches presented here, especially in connecting tau aggregation with nuclear speckle biology and RNA processing, will be of broad interest and offers important new mechanistic insights into tau pathology. I am an expert in Alzheimer's disease and integrated stress response.

  2. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    In their manuscript, Decker et al., investigate the relationship between cytoplasmic and nuclear tau aggregation using a HEK293T biosensor system and propose a multistep model in which cytoplasmic aggregates give rise to nuclear tau aggregates, potentially via VCP-dependent secondary seed generation and involvement of nuclear speckle components. The study further explores functional consequences of nuclear tau aggregation on RNA metabolism. Overall, the work is interesting and potentially impactful. The combination of live-cell imaging, seed-labeling strategies, genetic perturbations (SRRM2/PNN), and RNA imaging represents a thoughtful experimental approach. However, I have some minor concerns and feel the authors should address these -

    1. Poly(A)+ FISH intensity is not a direct measure of export efficiency. The authors claim that tau aggregation within nuclear speckles interferes with nuclear export of RNA. It is highly possible that increased nuclear RNA levels observed could reflect altered transcription, stability, or stress responses rather than export defects alone. In the case of ATF3, a known stress responsive gene, increased nuclear signal could reflect transcriptional activation, not export defects. To prove that export is defective, the authors should at least measure total RNA levels (qPCR) in nuclear vs cytoplasmic fraction.
    2. Though the authors have shown the proposed role of VCP in generating secondary seeds by using inhibitors, the authors should show genetic validation by using dominant-negative VCP.

    Referees cross-commenting

    This session contains comments from different reviewers

    Reviewer 3

    I agree with the reviewers that additional controls and experiments would strengthen the VCP inhibition studies. However, I would like to clarify that the specific concern raised by Reviewer 1 (Key point number 4) regarding fusion of tau to GFP does not apply to this manuscript. In this study, the authors use tau conjugated to Cy3, a well established approach in the field that adds only approximately 1 kDa to the protein.

    Reviewer 1

    Apologies reviewer 3, but I respectfully disagree. Please look again at the legends for figs 1 through fig 5. All clearly delineate the use of tau biosensor cells using a YFP rather than GFP fusion protein with tau. i do agree we should correct my review to state YFP rather than GFP, but structurally the concern remains the same. Cy3 labelling, I believe is used to track the relatively short lived exogenous seeds.

    Significance

    The integration of approaches presented here, especially in connecting tau aggregation with nuclear speckle biology and RNA processing, will be of broad interest and offers important new mechanistic insights into tau pathology. I am an expert in Alzheimer's disease and integrated stress response.

  3. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    The manuscript by Decker et al. examines the formation of nuclear tau aggregates and their functional consequences using a HEK293T tau biosensor system. The authors propose that nuclear tau aggregates arise through VCP dependent secondary seeding from cytoplasmic aggregates and that these nuclear aggregates impair RNA export. The study addresses an important and relatively unexplored aspect of tau biology. However, several conclusions extend beyond what the data directly supports, and several essential controls are missing.

    Major Comments

    • The introduction is generally clear and well organized. However, it would benefit from additional mechanistic context explaining how polyserine domains might promote tau aggregation and why this feature is biologically relevant.
    • The live cell imaging convincingly demonstrates the temporal sequence of cytoplasmic followed by nuclear aggregation. However, the manuscript lacks controls assessing whether nuclear aggregation correlates with higher tau expression levels. Such controls are necessary to exclude expression driven artifacts.
    • The authors conclude that nuclear envelope integrity is preserved, but only import assays were performed. To validate the sensitivity and specificity of the assay, export assays or positive controls for nuclear transport disruption are required.
    • The Cy3/Cy5 seed experiments support the claim that exogenous seeds do not enter the nucleus. However, the conclusion that VCP generates secondary seeds is overstated. For example, the manuscript states: "VCP is responsible for the formation of secondary seeds..." (lines 418-439), yet the data demonstrate correlation rather than direct evidence of seed generation. To substantiate this conclusion, the authors should: directly quantify seed abundance. The current interpretation assumes uniform cytoplasmic uptake of seeds but does not measure it; Include controls addressing VCP inhibitor specificity, as these compounds have pleiotropic effects (e.g., ER stress, proteostasis collapse). No data is provided on whether VCP inhibition alters tau ubiquitination, which could have major implications on tau aggregation.
    • The authors observed increase in nuclear poly(A)+ RNA and specific transcripts. However, the current data do not distinguish between several possible mechanisms that may account for this increase, including impaired export, increased transcription, enhanced RNA stability, or nuclear retention due to speckle reorganization.
    • The discussion occasionally overinterprets the data. Several statements should be reframed as hypotheses rather than conclusions:
    1. "VCP can generate tau seeds capable of additional seeding within a cell." (lines 572-594) This has not been directly demonstrated and should be softened accordingly.
    2. Active import via SRRM2 is proposed, but no supporting data are presented. This should be clearly framed as a speculative model.
    3. "Tau aggregates in the nucleus alter the function of nuclear speckles..." (lines 616-637). While plausible, this is not directly shown. Alternative explanations such as transcriptional upregulation or stress induced changes should be acknowledged.
    4. The statement "It is possible that such nuclear aggregates could alter nuclear RNA export and contribute to pathology." (lines 637-655) is reasonable, but the authors should emphasize that nuclear tau aggregates are not consistently observed across tauopathies and that the HEK293T biosensor system may not fully recapitulate neuronal biology.

    Significance

    The manuscript by Decker et al. examines the formation of nuclear tau aggregates and their functional consequences using a HEK293T tau biosensor system. The authors propose that nuclear tau aggregates arise through VCP dependent secondary seeding from cytoplasmic aggregates and that these nuclear aggregates impair RNA export. The study addresses an important and relatively unexplored aspect of tau biology. However, several conclusions extend beyond what the data directly supports, and several essential controls are missing.

  4. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    Decker et al present an interesting study of the order of events in tau seeding in a biosensor HEK293 derived cell line. This a critical unresolved question in the field about subcellular compartment contributions to tau aggregation. This exploration of the nuclear tau aggregated deposition and seeding in HEK293T tau biosensor cells uses a variety of imaging-based methods. They show that nuclear aggregates only form in cells with cytosolic aggregates, nuclear aggregates cannot form in the absence of cytoplasmic tau aggregates. The original tau seeds do not persist. Also aggregates in the nucleus are dependent on VCP and SRRM2 for formation. The paper is limited in scope by use of only HEK239T cells and seem to overstate the generalizability of the findings to neuronal mechanisms of neurodegeneration. Please see to all tauopathies. In particular, the manuscript does not outline the overstatement of some of the conclusions.

    Key Points to address:

    1. The manuscript does not detail limitations of the study in the discussion. Please address the concern that HEK293T biosensor cells are not neurons. Especially in the clear animations showing the transformation from cytoplasmic to nuclear aggregates appears to require cell division and nuclear breakdown.
    2. The introduction sets this up as Alzheimer's disease relevant but all studies are down with P301S tau which is a distinct and particularly aggressive form of tauopathy (FTLD-Tau). There is no amyloid beta component to any of these studies.
    3. The study does not address the peculiar structure of P301S aggregates, which while disease relevant are clearly distinct from AD or most forms of familial FTLD. The authors should limit the generalizability of the findings to their particular form of tauopathy unless they plan to use multiple tau fibril conformations in their studies.
    4. The authors do not address the potential impact of fusing a natively unfolded protein like tau to a highly structured beta barrel like GFP. Please present this potential confound.
    5. Inhibition of VCP can cause proteinopathies in the absence of other seeding. For instance, familial mutations in human VCP can cause either tau or TDP-43 proteinopathy depending on the specific human disease causing mutation. Thus, critical controls are missing from figure 3. For instance, the consequence of VCP inhibition on unseeded biosensor cells is a missing control. Second all panels should evaluate TDP-43 aggregation to ascertain whether or not the secondary nuclear seeding involves TDP-43.

    Minor concerns:

    A. Line 635 - In line 380, they discuss that aggregation of tau does not lead to perturbations in nuclear transport. In line 390, they discuss that aggregation of tau does not affect nuclear envelope integrity or nuclear import. However, in the discussion discusses that aggregation alters nuclear RNA export. These statements could use clarifying that protein export is not perturbed but RNA export and import may be.

    B. Line 564: "This observation suggests that tau aggregation in the cytoplasm may lead to increased expression of some RNAs." This could also be that cytoplasmic tau alters RNA export. These experiments don't differentiate between these options.

    C. In Figure 1, the authors show large aggregates overlapping the nucleus. It is unclear whether these aggregates have a portion both within and outside the nucleus or if they are deforming the nucleus and are wholly external to the nuclear compartment. Clarity on this issue is important. If the nucleus is deferment the observed aggregates seem reminiscent of aggresome formation. Please clarify.

    Significance

    Decker et al present an interesting study of the order of events in tau seeding in a biosensor HEK293 derived cell line. This a critical unresolved question in the field about subcellular compartment contributions to tau aggregation. This exploration of the nuclear tau aggregated deposition and seeding in HEK293T tau biosensor cells uses a variety of imaging-based methods. They show that nuclear aggregates only form in cells with cytosolic aggregates, nuclear aggregates cannot form in the absence of cytoplasmic tau aggregates.

  5. Version published to 10.64898/2026.02.17.706203 on bioRxiv