8 July 2010. Large proteins need a key—in the form of a nuclear localization sequence—to unlock the transporters that let them into the cell’s inner sanctum. Without the right key, fused in sarcoma (FUS) gets left outside in the cytoplasm, wreaking havoc that can cause neurodegeneration, according to a paper released yesterday by the EMBO Journal. The authors, from the German Center for Neurodegenerative Diseases and the Ludwig-Maximilians-University in Munich, propose that the protein lockout is the first step toward FUSopathies, including amyotrophic lateral sclerosis and frontotemporal lobar dementia. Senior author Christian Haass will present these results at the International Conference on Alzheimer’s Disease in Honolulu, Hawaii, on July 14. The study is the first to ascribe function to the region of FUS harboring some familial ALS and FTLD mutations, according to an accompanying commentary by Emanuele Buratti and Francisco Baralle of the International Center for Genetic Engineering and Biotechnology in Trieste, Italy.
FUS is involved in RNA processing in the nucleus. Mutations somehow cause the protein to accumulate in cytoplasmic aggregates and lead to several varieties of neurodegeneration (see ARF related news story). FUS is similar to TDP-43, another nuclear protein linked to ALS, which shares an RNA-processing function and accumulates in degenerating neurons as well. These commonalities suggest the two might cause disease in the same manner (see ARF related news story on Kwiatkowski et al., 2009 and Vance et al., 2009).
First author Dorothee Dormann and colleagues reasoned that the simplest explanation for cytoplasmic FUS accumulation would be a mutation in the nuclear localization sequence (NLS). However, FUS does not contain a classical NLS. What it does have is several arginines, a proline and a tyrosine at the carboxyl-terminus, right where 12 of the 20 known disease-causing FUS mutations cluster. These amino acids are characteristic of a non-classical, proline-tyrosine (PY)-NLS, a variable signaling motif of 12 to 15 amino acids that cannot be confirmed by simple sequence-gazing (Lee et al., 2006).
To confirm or deny the presence of a PY-NLS in FUS, the researchers deleted or altered its carboxyl-terminal domain and expressed the mutants in HeLa cells. While wild-type FUS targeted the nucleus, the mutants were more likely to be found in the cytoplasm, showing the carboxyl-terminal domain is necessary for nuclear import. To show the domain is also sufficient for nuclear localization, the scientists joined it to a green fluorescent protein-glutathione S-transferase chimera (GFP-GST), which is too big to gain nuclear entry on its own. With the last 13 amino acids of FUS attached, GFP-GST did go nuclear.
Next, Dormann and colleagues examined localization of the FUS mutations known to cause disease. These abandoned the nucleus to produce diffuse cytoplasmic staining. These mutations cause varying phenotypes, with the most severe causing disease onset at approximately 24 years, and the weakest having incomplete penetrance with some carriers living disease-free into their forties and beyond. The degree of cytoplasmic mislocalization in cell culture was proportional to mutation severity. “A weak mutation shows very little, if any, redistribution,” Haass said. “With the much stronger mutations, you get tremendous cytoplasmic localization.” But even the worst mutations left up to half of the cell’s FUS in the nucleus. That amount is likely enough for neurons to survive, as long as no other stress assaults the cell, suggested Haass.
Could stress, then, be the second ingredient in FUSopathy? To find out, the researchers stressed the cells by preventing multiple PY-NLS-dependent proteins from entering the nucleus. Non-classical PY-NLSs use a protein called transportin as their portal. A peptide called M9M blocks the door, not only for FUS but also for the other proteins that rely on this gateway. Therefore, M9M is likely to stress the cell by preventing several proteins from accessing the nucleus.
When the researchers treated both HeLa cells and primary neurons with M9M, they observed diffuse FUS staining as well as cytoplasmic punctate structures. They identified these structures as stress granules, protein complexes that protect mRNA during stress. Disease-linked FUS mutants, normally evenly distributed in the cytoplasm, also appeared in stress granules when cells were heat-shocked. Dormann and colleagues concluded that FUS, when shut out of the nucleus, remains diffuse unless the cell experiences additional stress.
A similar process appears to occur in human FUSopathies. The researchers collected postmortem neural samples from people who had different FUS-based diseases, including some cases where FUS is wild-type. In all cases, FUS aggregates included markers for stress granules. The aggregates in the tissue samples were also bigger than those in cultured cells.
Haass and colleagues propose a double-hit hypothesis for familial FUS diseases. The first hit is the mutation itself, which minimizes the amount of FUS that can enter the nucleus. However, some FUS is still able to reach the nucleus, so there is insufficient loss of function to cause sickness. The second hit occurs when some environmental stress—perhaps the natural effects of aging—spurs stress granule formation in the cytoplasm. Then FUS joins the granules, recruiting more FUS in the process and keeping it from the nucleus. This suggests that the less FUS a person has in the nucleus to start with, the earlier stress granules collect enough FUS to cause symptoms by preventing its nuclear localization and function—whatever that may be. The next step, Haass suggested, would be to determine what RNAs require FUS’s assistance in the nucleus.
The research does not yet explain how wild-type FUS might be associated with disease. The majority of ALS cases arise sporadically, and are not inherited, although some sporadic cases do have FUS mutations (Chiò et al., 2010). The FUS mutations cause the most severe form of disease, Haass said, but something similar yet milder may occur with normal FUS. Aggregated FUS does appear in sporadic ALS (see ARF related news story on Deng et al., 2010; see also Matsuoka et al., 2010). In addition, there is still an open question about how amino-terminal FUS mutations interfere with the protein. Dormann and colleagues found no evidence that these mutations affect nuclear transport. Perhaps they cause aggregation by other means, suggested Aaron Gitler of the University of Pennsylvania in Philadelphia, who was not involved in the study.
The results are somewhat surprising, Gitler said, because scientists had suspected that FUS and TDP-43 mutations cause disease in similar ways. TDP-43 possesses a classical NLS, but thus far researchers have not found any mutations in this region in people with neurodegenerative disease. However, Gitler said, some TDP-43 mutations do increase protein aggregation. The current work suggests at least some FUS mutations have a different mechanism. “Maybe these proteins are not so similar, after all,” Gitler said.—Amber Dance.
Dormann D, Rodde R, Edbauer D, Bentmann E, Fischer I, Hruscha A, Than ME, Mackenzie IRA, Capell A, Schmid B, Neumann M, Haass C. ALS-associated fused in sarcoma (FUS) mutations disrupt Transportin-mediated nuclear import. EMBO J. 2010 Jul 7. Abstract
Buratti E, Baralle FE. Neurons don’t appreciate FUSsing in the cytoplasm. EMBO J. 2010 Jul 7. Abstract