A single point mutation in the fused in sarcoma (FUS) gene can throw a wrench into the delicate mechanics of protecting and splicing genetic material, according to a new study. Reporting in the February 10 Journal of Clinical Investigation, researchers led by Eric Huang at the University of California, San Francisco, debut a new transgenic mouse model that expresses the most common FUS mutation found in patients with familial amyotrophic lateral sclerosis (fALS). Their findings add to a thicket of data collected from prior models of FUS.

“We have learned from this animal model that even if we just look at a single point mutation, the effects are complicated,” Huang, the senior author, told Alzforum. 

First author Haiyan Qiu and colleagues found that mice expressing human FUS with an arginine-to-cysteine mutation at residue 521 (R521C) displayed profound motor defects and died young. Their surviving motor neurons harbored damaged DNA and improperly spliced RNA. The work is impressive, and shows that the FUS mutation “touches all aspects of gene expression,” said Thomas Kukar of Emory University in Atlanta, who was not involved in the study. 

Multiple mutations in FUS cause familial ALS. Insoluble cellular inclusions containing the protein are a hallmark of this disease and the related frontotemporal lobular dementia (see Feb 2009 news story). FUS plays a role in DNA repair and RNA splicing in the nucleus, and participates in shuttling RNA into the cytoplasm. 

Like most FUS mutations identified so far, R521C occurs at the C-terminus of the protein and disrupts a motif needed for entry into the nucleus (see Vance et al., 2013). Previous models based on overexpression of FUS showed accumulation of the protein in the cytoplasm of neurons (see Mitchell et al., 2013, and Verbeeck et al., 2012). Huang wanted to generate a model more akin to human disease, where the mutant protein is expressed at levels similar to the endogenous protein. 

The researchers generated mice expressing FUS-R521C under control of the Syrian hamster prion promoter, and western blots of the brain and spinal cord indicated that the mutant protein was expressed at similar levels as the endogenous protein. The mice displayed severe ALS symptoms within the first few weeks of life and died shortly thereafter. Huang presented the initial results at a Keystone Symposium last year (see Feb 2013 conference coverage). As Alzforum reported, Huang said that the mice retained half of their motor neurons despite their severe disease. Those surviving motor neurons had primitive dendrite branches and abnormally few synapses. He also found that mutant FUS formed stable complexes with itself and with the native mouse protein, suggesting that the mutant kept normal FUS from maintaining nucleic acids. Likely due to these complexes, both the mutant and wild-type proteins in the spinal cord of transgenic mice accumulated two- to threefold over non-transgenic mice. Now, the researchers report that the FUS mutation leads to faulty DNA repair and gene splicing as well, and that, in turn, dysregulates gene transcription.   

Jibing with what Huang’s group had previously seen in vitro, the neurons from the mutant mice had DNA damage, including characteristic comet tail-shaped nuclei. To get a sense for which neuronal genes were most susceptible to DNA damage, the researchers screened for oxidized purine residues. They discovered that the gene for brain-derived neurotrophic factor (BDNF), a key player in dendritic growth and synapse function, was peppered with damage.  

The researchers then scrutinized BDNF. This gene packs eight exons, which are differentially spliced, before the coding sequence. Because this process is complex and tightly regulated, the researchers tested whether mutant FUS disrupted it. Using cross-linking and co-immunoprecipitation (CLIP), the investigators found that mutant FUS, more than wild-type, associated with certain splice junctions, suggesting that the sticky mutant could hinder processing of mRNA. Consistent with this idea, neurons in mutant FUS mice made less BDNF than controls did, though the authors did not look for a change in splicing patterns.

Other genes may be affected in a similar fashion. Adding exogenous BDNF only partially rescued dendritic branching defects in cultured cortical neurons transfected with mutant FUS. In fact, the researchers identified 766 genes that were differentially expressed between the FUS mutant and control mice involved in synaptic function, others in immune response or DNA repair. In general, mutant mice retained more introns, suggesting widespread splicing defects. Given that post-mitotic neurons are highly sensitive to DNA damage, they may be particularly vulnerable when FUS goes wrong, said Huang. 

Researchers have struggled to create FUS-based mouse models that recapitulate classic pathologies associated with ALS. The early death of Huang’s mice, and the incomplete loss of motor neurons, are shared in mice developed in Chris Shaw’s lab at King’s College London (see Mitchell et al., 2013). That both models retain motor neurons—as opposed to SOD1 mutant mice, which lose more than 90 percent of theirs—is good news to Huang. “I look at this from the bright side,” he said. “The main problem is how these neurons connect with other neurons, so perhaps we can design a clever way to help them rewire, to help these patients.” Mice models of Alzheimer's also have little neuronal loss, though that is a major hallmark of the disease. 

The mutant FUS protein in Huang’s mice turned up primarily in the nucleus. This contrasts the cytoplasmic inclusions seen in both human disease and in Shaw's mice. Huang predicts that the motor neurons in his model have little time to amass the protein because the mice die so young. Whether the mechanisms of pathology are similar in the two models remains to be determined.

“We believe that the mutant protein causes most of its damage in the nucleus, where it binds the wild-type protein, interacts with the RNA splicing machinery, and also with the DNA damage repair process,” said Huang. A combination of loss and gain of function may cause disease, as the mutant protein may prevent the native protein from carrying out its normal functions, but also binds more tightly to splice junctions to directly foul up gene expression. 

Kukar’s lab developed an entirely different model in 2012, when the researchers injected adeno-associated viruses expressing mutant FUS proteins directly into the brains of mice (see Verbeeck et al., 2012). Kukar’s mice did accumulate FUS in the cytoplasm, especially the FUS-D14 mutant, which lacks the entire nuclear localization sequence. However, those mice did not show signs of motor defects or succumb to disease. Kukar attributes this difference to the fact that the virally delivered mutant is less widely expressed than in transgenic models. 

Kukar said that Huang’s mouse model was useful in pinpointing possible mechanisms of disease, but noted that some aspects of the DNA damage response differed between mice and humans. “If faulty DNA damage repair is playing an important role, which I suspect it is, then we probably need to also look in models that better reflect human disease,” Kukar said. He has high hopes that human induced pluripotent stem cells will yield informative models.

That Huang’s mice die so young contrasts to the human form of the disease. However, Udai Pandey at the University of Pittsburgh, who was not involved in the research, noted that strong phenotypes make helpful tools. “This was a needed model,” Pandey said. “It will be useful to further understand molecular mechanisms and identify new targets.”

Huang plans to use the FUS-R521C mouse model to understand how the mutant affects the assembly of splicing machinery, and to compare the effects to those of mutations in TDP-43, a protein also involved in RNA splicing, and C9ORF72.

Both Huang and Kukar noted that a knock-in mouse, in which endogenous mouse FUS is replaced by mutant FUS, could more accurately replicate endogenous expression levels in human disease. Such a model is in the works.—Jessica Shugart


  1. It is always encouraging when another model of ALS is published. The more models we have, the more we can find out and the faster we can get to therapies. This model specifically overexpresses a mutant form of FUS. The onset of disease occurs very early in these mice, which might relate directly to the overall levels of FUS expressed in these animals. Although the mutant form is expressed at a similar level to the endogenous one in these transgenics, the overall increase in expression is significantly higher compared to a non-transgenic. It has been shown in both FUS and TDP models of disease that the age of onset seems related to the level of overexpression rather than the mutation (both wild-type FUS and wild-type TDP overexpression mouse models show neurodegenerative-like phenotypes).

    It is also interesting that in these animals there is no sign of downregulation of the endogenous protein, which might result from the mutation because FUS is known to self-regulate (by splicing its own mRNA, which leads it to be degraded) and does so in other models (Mitchell et al., 2013). What is not clear is whether this model can tell us anything about FTLD-FUS. It might be that FTLD-FUS is not caused primarily by a FUS deficit but by a disruption of pathways in which it is involved. Certainly there are other proteins, such as TAF15 and EWS and TNPO1, which are found in FTLD-FUS inclusions but are not found in ALS-FUS inclusions.  

    The authors report that the mutant and wild-type FUS form complexes. That agrees with a variety of papers that suggest that this is the case. It is perhaps surprising that they did not find an interaction between the WT protein and itself, as has been previously published. I think that the formation of these stable complexes is very interesting, as it fits with the hypothesis that the mutant protein can sequester the wild-type protein within cytoplasmic aggregates in ALS. Given that a single base pair change can lead to such a devastating disease, it is not surprising that it might have such an effect in cell and animal models. 

    In this case I suspect that the lack of accumulation of FUS in the cytoplasm has more to do with longevity, because the level of FUS seems fairly toxic to the mice and they only live a short while. I suspect that the expression is high in these animals and that a lower sustained expression of FUS would more likely result in aggregates. Previous models showed that increasing the overall expression level of WT FUS by 1.4- to 1.9-fold was sufficient to result in an aggressive, fatal phenotype. I feel that this model, like most models, tells us something about the mechanism of disease but does not fully recapitulate the disease. After all, we are comparing a disease that takes a few decades to manifest (even in the FUS families) to a very rapid onset in the models. I think that all overexpression models, regardless of whether they are WT or mutant forms of the protein, tell us that FUS, like TDP43, is a heavily regulated protein and that under- or overexpression of it results in cell death. This suggests that there are critical cell functions that are carried out by FUS for which any disruption results in problems for the cell. 

    I think that the RNA work that has been done in this paper and the neuronal changes seen here, linked with recent publications on both TDP and FUS in the nervous system, suggests that a disruption of RNA metabolism in the neurons leading to altered synaptic function and ultimately cell death is a likely disease mechanism. 


    . Overexpression of human wild-type FUS causes progressive motor neuron degeneration in an age- and dose-dependent fashion. Acta Neuropathol. 2012 Sep 9; PubMed.

    View all comments by Caroline Vance

Make a Comment

To make a comment you must login or register.


Alzpedia Citations

  1. FUS

News Citations

  1. New Gene for ALS: RNA Regulation May Be Common Culprit
  2. Up-and-Coming ALS Mice Leave Scientists ConFUSed

Paper Citations

  1. . ALS mutant FUS disrupts nuclear localization and sequesters wild-type FUS within cytoplasmic stress granules. Hum Mol Genet. 2013 Jul 1;22(13):2676-88. Epub 2013 Mar 7 PubMed.
  2. . Overexpression of human wild-type FUS causes progressive motor neuron degeneration in an age- and dose-dependent fashion. Acta Neuropathol. 2012 Sep 9; PubMed.
  3. . Expression of Fused in sarcoma mutations in mice recapitulates the neuropathology of FUS proteinopathies and provides insight into disease pathogenesis. Mol Neurodegener. 2012;7:53. PubMed.

Further Reading


  1. . Converging mechanisms in ALS and FTD: disrupted RNA and protein homeostasis. Neuron. 2013 Aug 7;79(3):416-38. PubMed.
  2. . Fused in Sarcoma (FUS) Protein Lacking Nuclear Localization Signal (NLS) and Major RNA Binding Motifs Triggers Proteinopathy and Severe Motor Phenotype in Transgenic Mice. J Biol Chem. 2013 Aug 30;288(35):25266-74. PubMed.

Primary Papers

  1. . ALS-associated mutation FUS-R521C causes DNA damage and RNA splicing defects. J Clin Invest. 2014 Mar 3;124(3):981-99. Epub 2014 Feb 10 PubMed.