Some people with amyotrophic lateral sclerosis or frontotemporal dementia accumulate deposits of the nuclear protein FUS in the cytoplasm. What drives this relocation? In the July 24 Cell Reports, researchers led by Magdalini Polymenidou at the University of Zurich blame osmotic pressure. When Polymenidou and colleagues subjected brain slices to hyperosmotic conditions, FUS piled up in cytoplasm of neurons. The protein localized to stress granules, but even when the authors prevented those from forming, FUS still loitered in the cell body. The authors traced the cause of this mislocalization to transportin 1 and 2, the nuclear import factors that ferry FUS. These factors became marooned in cytoplasm during osmotic stress, they found.

  • Under hyperosmotic stress, FUS builds up in cytoplasmic stress granules in neurons.
  • Mislocalization of transportins, which escort FUS into the nucleus, may be to blame.
  • The findings suggest one way wild-type FUS might be prone to aggregate in FTD, ALS.

FUS mislocalization could be the first step toward disease, Polymenidou suggested. She believes osmotic stress acts as the trigger that sends FUS to the wrong place, where a second stressor could then cause it to aggregate. Supporting the disease connection, hyperosmolarity had no effect on FUS in astrocytes. These cells do not develop FUS deposits in ALS and FTD.

“It is a really exciting observation that the relocalization of FUS is independent of stress granule formation,” Brian Freibaum at St. Jude Children’s Research Hospital in Memphis, Tennessee, wrote to Alzforum (full comment below). Others were more cautious. “The potential contribution of this mechanism to ALS/FTD pathogenesis is unclear,” noted Sami Barmada of the University of Michigan Medical School in Ann Arbor (full comment below).

Transportins Control FUS.

Transportins, aka TNPO (green), escort FUS (red) into the nucleus (top), but under hyperosmotic conditions (left), both get stuck in the cytoplasm. Astrocytes are unaffected (right). [Courtesy of Cell Reports, Hock et al.]

FUS contains a nuclear localization signal that allows transportins 1 and 2 (collectively called TNPO) to carry it into the nucleus. Without TNPO, FUS that escapes the nucleus via passive diffusion gradually amasses in cytoplasm (May 2018 news). Even during most types of cellular stress, including oxidative stress, ER stress, and heat shock, TNPO keeps FUS corralled in the nucleus. One previous report flagged osmotic stress as the exception. Researchers led by Daryl Bosco at the University of Massachusetts Medical School found that subjecting immortalized somatic cell lines to hyperosmolarity forced FUS into stress granules (Sama et al., 2013). 

Polymenidou and colleagues wondered if the same thing would happen in an environment closer to that of living brain. First author Eva-Maria Hock applied sorbitol, a sugar alcohol, to mouse cortico-hippocampal slices to induce osmotic stress. The high concentration of solutes in the extracellular fluid pulls water out of cells in the slices, shrinking them and causing intracellular molecular crowding. After two hours of this treatment, FUS began to disappear from nuclei, showing up in cytoplasmic stress granules. This occurred in neurons and microglia, but not astrocytes. Cytoplasmic FUS peaked at four hours. Then, curiously, it began to shift back to its normal location. After eight hours of continuous osmotic stress, FUS was confined to the nucleus again, and stress granules had dissolved, suggesting cells somehow adapted to these conditions. As with the cell lines, FUS mislocalization occurred only under osmotic stress; treating slices with arsenite to induce oxidative stress did not budge the protein.

Did FUS movement depend on stress granule formation? To test this idea, the authors treated the slices with inhibitors of protein synthesis, which prevent stress granule formation. FUS still migrated to cytoplasm under osmotic stress, although it remained diffuse (see image below). Likewise, artificially stabilizing stress granules did not trap FUS in cytoplasm; it still escaped back to the nucleus after eight hours.

Osmosis Moves FUS.

FUS (red) stays nuclear under normal conditions (left), but enters cytoplasmic stress granules (green; overlay appears yellow) during hyperosmotic stress (middle). Preventing stress granule formation results in diffuse cytoplasmic FUS (right). Nuclei are blue. [Courtesy of Cell Reports, Hock et al.]

“It surprised me that translocation [of FUS] to cytoplasm and incorporation into stress granules were independent processes. That’s an important conceptual change in how we think about these steps,” Polymenidou told Alzforum. Some evidence suggests this uncoupling might be specific to osmotic stress. Researchers at Johns Hopkins University in Baltimore recently reported that stress granules formed by oxidative stress actively sequester nucleocytoplasmic transport factors like TNPO. In this case, inhibiting granule formation prevented mislocalization of these factors and blocked neurodegeneration (Zhang et al., 2018). 

The authors explored several possible mechanisms for this FUS mislocalization. Signaling pathways and transcription factors activated by hyperosmotic stress did not affect FUS movement. Turning to nuclear import, the authors saw that other TNPO cargos accumulated in the cytoplasm during osmotic stress as well, but cargos of other types of importins did not. Immunostaining revealed that TNPO fled the nucleus and localized to stress granules during osmotic stress, in contrast to its normal diffuse distribution throughout the nucleus and cytoplasm. Transfecting cells with a TNPO construct that trapped it in the nucleus kept FUS in the nucleus during hyperosmolarity.

Commenters were particularly interested in the TNPO connection. Wilfried Rossoll at the Mayo Clinic in Jacksonville, Florida, noted that several recent studies have shown a role for transportins in preventing RNA-binding proteins such as FUS from sticking together (Apr 2018 news). “Since hyperosmolarity causes cell shrinkage and macromolecular crowding, these new findings suggest that cytoplasmic recruitment of transportins may act as a protective mechanism to counteract increased levels of protein aggregation,” Rossoll wrote to Alzforum (full comment below).

Were the findings on FUS specific to mouse cells? The authors repeated the experiments in human neurons derived from stem cells, and saw the same response to osmotic stress. However, human neurons did not adapt as quickly or as well to this type of stress as mouse neurons did, with FUS remaining in the cytoplasm of some cells under osmolar conditions. By contrast, FUS in human astrocytes was unaffected, just as in mouse astrocytes, although the cells showed other signs of osmotic stress, such as the activation of signaling cascades.

“I was very surprised to see that FUS in astrocytes resisted cytoplasmic relocalization,” Freibaum said. Polymenidou said she is investigating the reason for this. She thinks the mechanism might point toward potential interventions that could help neurons deal better with osmotic stress. Keeping FUS nuclear might prevent pathogenic deposits from forming, she suggested. She believes these findings are particularly pertinent for FTD. In ALS, it is mutant FUS that gets trapped in the cell body, but in FTD, it is the wild-type protein. The cytoplasmic mislocalization of normal FUS was a mystery.

It is unclear, however, how common hyperosmolarity is in the brain, and whether this condition could precipitate disease. Hyperosmolarity is sometimes induced in brain after traumatic injuries to bring down swelling; Polymenidou speculated that this could be one reason for the association between TBI and neurodegeneration. Severe dehydration or an imbalance of electrolytes might also cause osmotic stress in the brain, she said.

Bosco noted that the experimental conditions used in this study, like those in in her own, are not physiological. However, the molecular crowding caused by hyperosmotic stress might mimic other conditions that occur in the brain, such as impaired proteostasis and protein accumulation, she suggested. She also noted that hyperosmolarity can trigger the release of proinflammatory cytokines (Schwartz et al., 2009; Brocker et al., 2012). Inflammation is a known risk factor for many neurodegenerative diseases.—Madolyn Bowman Rogers


  1. It is really exciting to observe that the relocalization of FUS is independent of stress granule formation. This suggests a very unique role for TNPO transportins as well, since their relocalization to the cytoplasm is also independent of stress granule localization. I was also very surprised to see that FUS in astrocytes was resistant to cytoplasmic relocalization.

    I believe that it is possible osmotic stress plays a role in initiating FUS aggregation leading to human disease, but it’s also possible that aggregation of FUS in disease may be more influenced by genetic factors or through unknown stress mechanisms. In either case, osmotic stress looks like a valuable tool to understand the progression of FUS aggregation in human disease.

  2. This manuscript touches on a central theme in ALS, FTD, and related neurodegenerative disorders—the redistribution of RNA binding proteins in association with neurodegeneration. The trigger for cytoplasmic accumulation of these proteins, including TDP43 and FUS, remains unknown. Here, the authors show that hyperosmolar stress leads to mislocalization of FUS but not TDP43, and this seems to depend on redistribution of transportins to the cytoplasm. In opposition to a recent study (Zhang et al., 2018) suggesting that key nucleocytoplasmic transport factors are sequestered by stress granules, Hock et al. demonstrate that FUS mislocalization with hyperosmolar stress is independent of stress granules themselves. Even more interesting is the fact that this type of FUS mislocalization is unique to hyperosmolar stress, and is not witnessed with other forms of cellular stress—this is not a conserved pathway, but rather specific. Moreover, the phenomenon is absent in astrocytes, although readily apparent in neurons and in microglia.

    The potential contribution of this mechanism to ALS/FTD pathogenesis is unclear, however. While the authors were unable to detect FUS cytoplasmic inclusions in astrocytes, that does not mean that astrocytes do not play an essential role in the onset and/or progression of disease. Several lines of evidence indicate that astrocytes actively contribute to neurodegeneration in ALS, and perhaps also in FTD (see, for example Song et al., 2016; and Re et al., 2014). Additionally, while the two-hit hypothesis for FUS or TDP43 mislocalization is appealing, it remains to be seen if hyperosmolar stress in particular could be one such “hit.” The authors propose that hyperosmolar therapy during acute treatment for traumatic brain injury may be at least partially responsible for long-term changes in the risk of dementia. One could imagine testing this hypothesis by identifying and following those who received hyperosmolar therapy versus those who did not. 


    . Stress Granule Assembly Disrupts Nucleocytoplasmic Transport. Cell. 2018 May 3;173(4):958-971.e17. Epub 2018 Apr 5 PubMed.

    . Major histocompatibility complex class I molecules protect motor neurons from astrocyte-induced toxicity in amyotrophic lateral sclerosis. Nat Med. 2016 Apr;22(4):397-403. Epub 2016 Feb 29 PubMed.

    . Necroptosis drives motor neuron death in models of both sporadic and familial ALS. Neuron. 2014 Mar 5;81(5):1001-8. Epub 2014 Feb 6 PubMed.

  3. This is a very interesting and thought-provoking paper that sheds light on potential disease mechanisms for frontotemporal dementia (FTD) with FUS protein pathology. It demonstrates that osmotic stress-induced cytoplasmic localization of Transportin 1/2 nuclear import receptors causes mislocalization of FUS and other proteins that depend on transportins for nuclear translocation. While Transportin 1 has recently been identified as a novel stress granule (SG) protein (Markmiller et al., 2018) and SG assembly may disrupt nucleocytoplasmic transport (Zhang et al., 2018), this cytoplasmic accumulation of FUS and transportins appears to be independent of their recruitment into cytoplasmic SGs. This is especially intriguing in the light of related findings recently published back-to-back in Cell, that demonstrate an additional role for Transportin 1 as a chaperone that prevents pathological aggregation of FUS (Guo et al., 2018; Hofweber et al., 2018; Qamar et al., 2018; Yoshizawa et al., 2018). Since hyperosmolarity causes cell shrinkage and macromolecular crowding, these new findings suggest that cytoplasmic recruitment of transportins may act as a protective mechanism to counteract increased levels of protein aggregation, but also cause further cytoplasmic redistribution of aggregation-prone nuclear RNA-binding proteins.

    While hypertonic stress in vitro mimics some of the patterns of pathology seen in FTD, additional factors are likely required to cause disease-specific aggregate formation observed in these patients. It will be interesting to see whether findings from this study can be translated into animal models of FTD, e.g., via intracranial sorbitol infusions.​


    . Context-Dependent and Disease-Specific Diversity in Protein Interactions within Stress Granules. Cell. 2018 Jan 25;172(3):590-604.e13. PubMed.

    . Stress Granule Assembly Disrupts Nucleocytoplasmic Transport. Cell. 2018 May 3;173(4):958-971.e17. Epub 2018 Apr 5 PubMed.

    . Nuclear-Import Receptors Reverse Aberrant Phase Transitions of RNA-Binding Proteins with Prion-like Domains. Cell. 2018 Apr 19;173(3):677-692.e20. PubMed.

    . Phase Separation of FUS Is Suppressed by Its Nuclear Import Receptor and Arginine Methylation. Cell. 2018 Apr 19;173(3):706-719.e13. PubMed.

    . FUS Phase Separation Is Modulated by a Molecular Chaperone and Methylation of Arginine Cation-π Interactions. Cell. 2018 Apr 19;173(3):720-734.e15. PubMed.

    . Nuclear Import Receptor Inhibits Phase Separation of FUS through Binding to Multiple Sites. Cell. 2018 Apr 19;173(3):693-705.e22. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. No, TDP-43 and FUS Are Not Actively Exported From the Nucleus
  2. Liquid Phase Transition: A Deluge of Data Points to Multiple Regulators

Paper Citations

  1. . FUS/TLS assembles into stress granules and is a prosurvival factor during hyperosmolar stress. J Cell Physiol. 2013 Nov;228(11):2222-31. PubMed.
  2. . Stress Granule Assembly Disrupts Nucleocytoplasmic Transport. Cell. 2018 May 3;173(4):958-971.e17. Epub 2018 Apr 5 PubMed.
  3. . Is inflammation a consequence of extracellular hyperosmolarity?. J Inflamm (Lond). 2009 Jun 23;6:21. PubMed.
  4. . The role of hyperosmotic stress in inflammation and disease. Biomol Concepts. 2012 Aug;3(4):345-364. PubMed.

Further Reading

Primary Papers

  1. . Hypertonic Stress Causes Cytoplasmic Translocation of Neuronal, but Not Astrocytic, FUS due to Impaired Transportin Function. Cell Rep. 2018 Jul 24;24(4):987-1000.e7. PubMed.