Panel of Mouse Models Agree: Mutated FUS Gains Toxic Function

Mutations in FUS cause amyotrophic lateral sclerosis, but do they do that by interfering with the protein’s normal actions, engendering new toxic ones, or both? According to a paper in the February 4 Nature Communications, mutated FUS proteins gain toxic function. When knocked into mice, two different mutant versions of human FUS crippled motor neurons, whereas knocking in the wild-type human protein left the neurons unscathed. In contrast, knocking out mouse FUS caused no problem at all. The knock-ins provide a new disease model for the field, said senior author Neil Shneider of Columbia University in New York. Though they live a normal lifespan, their motor neurons start to degenerate by a few months of age, giving scientists a straightforward readout as they test treatments that might speed up or slow disease.

“These mice will be very helpful in figuring out the molecular mechanism and cell biology underlying ALS,” agreed Simon Alberti of the Max Planck Institute of Molecular Cell Biology and Genetics in Dresden, Germany, who did not participate in the study (see full comment below). 

Disappearing act.

Motor neurons (green) are sparser in spinal cords of FUS mutant mice (right) than in spinal cords expressing wild-type human FUS (left). [Courtesy of Sharma et al., Nature Communications, Creative Commons License.]

Scientists have been working to establish FUS models of ALS since the gene was first discovered (see Feb 2009 news; Jul 2009 conference news; Feb 2013 conference news). Many research groups started by overexpressing the protein, and they found that even wild-type FUS expressed at high levels can cause neurodegeneration, paralysis, and premature death (Mitchell et al., 2012). That, however, does not match human disease, Shneider noted. “You do not see supraphysiological levels of FUS in patients,” he said.

Co-first authors Aarti Sharma and Alexander Lyashchenko aimed for a more physiological model by knocking a single copy of human FUS into otherwise normal mice. The authors made three strains, inserting wild-type FUS, the R521C version associated with adult-onset ALS, or the P525L variant linked to juvenile-onset disease into the tau locus. Shneider had previous success using that locus for knock-ins, he said. Would that not interfere with tau physiology? Shneider told Alzforum that the remaining mouse tau gene compensates, producing more tau RNA to keep tau protein levels normal. Deleting one copy of tau did not noticeably change the animals’ phenotype, he added, nor did the FUS transgenes affect levels of tau or endogenous mouse FUS.

Overall, mice expressing wild-type or mutant human FUS were fairly healthy. The ones with FUS-P525L weakened over time, but subtly. When year-old mice dangled from a wire by their hind legs, FUS-P525L mice let go after about 40 seconds, compared to 80 seconds for the FUS-WT strain. While the authors did not report grip strength for older mice, all the animals lived a normal lifespan. “It is a modest phenotype,” Shneider admitted.

Nevertheless, the authors found a progressive neurodegeneration that began within months. In the FUS-P525L animals, motor neurons started to disappear by the time the mice were one month old. For the FUS-R521C mice, degeneration started by two months. By the time the mice were 12 months of age, the authors calculated a 30 percent loss of lumbar motor neurons in P525L mice, and 20 percent loss in R521C mice, compared to non-transgenic littermates. In the transgenics expressing wild-type human FUS, the motor neurons also remained healthy over two years. In addition, the authors created mouse lines in which the FUS transgene was only expressed in motor neurons; they observed similar mutation-dependent patterns of neurodegeneration in those animals.

Shneider was encouraged that the phenotype depended on there being a mutation, and that there were differences between FUS-R521C and FUS-P525L. “Just like in humans, the more aggressive allele is also more pathogenic in the mouse. This gave us some confidence that what we are seeing is disease-relevant,” he said.

The mice recapitulated another aspect of human disease, that is, the vulnerability of specific sets of motor neurons. In ALS, the motor neurons that innervate fast-twitch muscles are the first ones to pull back from neuromuscular junctions, then degenerate, while motor neurons that innervate slow-twitch muscles resist degeneration a while longer (see Jan 2015 news; Oct 2013 news; Mar 2009 news). The authors examined the neuromuscular junctions in hind leg fast-twitch tibialis anterior and slow-twitch soleus muscles, and found the knock-in mice replicated this pattern. In the tibialis, junctions started to dissociate by 20 days of age in the P525L mice, and by 40 days in the R521C. By a year, the P525L mice had lost 37 percent of the junctions in their tibialis compared with control mice, and the R521C mice were down by 30 percent. As for the soleus, the number of junctions stayed normal in the R521C mice out to a year, while the P525L mice lost about 10 percent. In mice with a wild-type FUS knock-in gene, neuromuscular junction numbers matched non-transgenic controls.

Despite these similarities to human motor neuron degeneration, the authors did find a difference between their mice and ALS when they examined where FUS was within neurons. FUS normally resides in the nucleus, but in diseased neurons and glia it relocates to the cytoplasm, where it forms inclusions. In knock-in motor neurons, mutant FUS indeed was in the cytosol and dendrites, while wild-type human FUS remained nuclear. However, despite this mislocalization, the authors found no evidence for the type of FUS inclusions that characterize ALS.

Nothing Lost, Something Gained
Based on these results, the authors could not tell if the neurodegeneration resulted from a gain of toxic function by the mutant FUS, or from loss of endogenous function due to the mutant interfering with normal mouse FUS. To test for the latter, they generated FUS knockouts. Because FUS-deficient mice do not survive past birth, the researchers engineered a strain in which they could snip out a portion of the gene at any time using the Cre/Lox gene-editing system. They activated Cre soon after birth, leading to the elimination of FUS protein by two weeks of age. The mice grew normally. “At one year, without FUS, motor neurons are happy as can be. There is no degeneration,” Shneider said. The same held true when they eliminated mouse FUS specifically in motor neurons. Knockout of endogenous FUS also made no difference to the various knock-in mice, indicating that total FUS levels did not affect the neurodegeneration.

Scientists who spoke with Alzforum said Shneider’s model mice offer important advantages. The single copy of FUS transgene may be more relevant to human FUS levels than multi-copy models, commented Tom Kukar of Emory University in Atlanta, who was not involved in the work. Scientists using overexpression models can never be sure if differences between, say, wild-type and mutant FUS are due to the FUS mutation or the copy number and location of the transgenes in the genome. “Of the FUS lines out there, it seems to be one of the cleaner ones, in the sense that the phenotypes are related to ALS-linked mutations,” Kukar said. For these reasons scientists have also turned to knock-in mice to study the effect of mutations in the amyloid precursor protein that causes Alzheimer’s disease (see Saito et al., 2014, and Alzforum webinar).

Larry Hayward of the University of Massachusetts Medical School in Worcester also praised the use of single-copy transgenics. However, he noted that the transgenes were expressed at different levels. Human FUS in the brain and spinal cord of the mutant knock-ins was about four times as concentrated as in the wild-type knock-ins, the authors found. They believe this is the case because the mutant protein is more stable. Hayward thought that the apparently more-toxic P525L FUS protein was even more concentrated than the R521C. He speculated that the amount of FUS protein present, in addition to its genotype, might influence the severity of neurodegeneration in the mice. “Expression level may be key,” Hayward said.

Shneider has offered to share these mice with other scientists, and has already sent them to several labs.—Amber Dance

Comments

  1. I think this is very nice work. It shows that FUS nuclear localization signal mutants may act through a toxic gain of function rather than a loss of function, which is fully consistent with the idea that aberrant phase transitions are the underlying mechanism. The defect seems to be qualitative rather than quantitative, which could be because of the specific recruitment of essential factors into aberrant granules that contain mutant FUS. Whether there really is no cross-seeding with wild-type FUS is difficult to tell from the mostly genetic data that are presented. The study points to perturbation of neuronal transport granules or local translation as the primary cause of the disease. These mice will be very helpful in figuring out the molecular mechanism and the cell biology underlying ALS.

    View all comments by Simon Alberti

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. New Gene for ALS: RNA Regulation May Be Common Culprit
  2. London, Ontario: The Fuss About FUS at ALS Meeting
  3. Up-and-Coming ALS Mice Leave Scientists ConFUSed
  4. Endoplasmic Reticulum Protein Protects Motor Neurons from ALS
  5. Surprise Save: Excitability Protects Neurons from Lou Gehrig’s
  6. ER Struggles in Motor Neurons That Fall to ALS

Webinar Citations

  1. Good-Bye Overexpression, Hello APP Knock-in. A Better Model?

Paper Citations

  1. Mitchell JC, McGoldrick P, Vance C, Hortobagyi T, Sreedharan J, Rogelj B, Tudor EL, Smith BN, Klasen C, Miller CC, Cooper JD, Greensmith L, Shaw CE. Overexpression of human wild-type FUS causes progressive motor neuron degeneration in an age- and dose-dependent fashion. Acta Neuropathol. 2012 Sep 9; PubMed.
  2. Saito T, Matsuba Y, Mihira N, Takano J, Nilsson P, Itohara S, Iwata N, Saido TC. Single App knock-in mouse models of Alzheimer's disease. Nat Neurosci. 2014 May;17(5):661-3. Epub 2014 Apr 13 PubMed.

Further Reading

Papers

  1. Sabatelli M, Moncada A, Conte A, Lattante S, Marangi G, Luigetti M, Lucchini M, Mirabella M, Romano A, Del Grande A, Bisogni G, Doronzio PN, Rossini PM, Zollino M. Mutations in the 3' untranslated region of FUS causing FUS overexpression are associated with amyotrophic lateral sclerosis. Hum Mol Genet. 2013 Jul 24; PubMed.
  2. Suzuki H, Matsuoka M. Overexpression of nuclear FUS induces neuronal cell death. Neuroscience. 2015 Feb 26;287:113-24. Epub 2014 Dec 12 PubMed.
  3. Ward CL, Boggio KJ, Johnson BN, Boyd JB, Douthwright S, Shaffer SA, Landers JE, Glicksman MA, Bosco DA. A loss of FUS/TLS function leads to impaired cellular proliferation. Cell Death Dis. 2014 Dec 11;5:e1572. PubMed.
  4. Haass C. Cellular mechanisms of ALS mutations - a loss or a gain of function?. Drug Res (Stuttg). 2013 Nov;63 Suppl 1:S17. Epub 2013 Nov 15 PubMed.

Primary Papers

  1. Sharma A, Lyashchenko AK, Lu L, Nasrabady SE, Elmaleh M, Mendelsohn M, Nemes A, Tapia JC, Mentis GZ, Shneider NA. ALS-associated mutant FUS induces selective motor neuron degeneration through toxic gain of function. Nat Commun. 2016 Feb 4;7:10465. PubMed.

Annotate

To make an annotation you must Login or Register.