16 July 2010. TDP-43 is intimately linked with amyotrophic lateral sclerosis, but the mystery remains as to what the protein is normally meant to do, and what goes wrong when it mutates. Two recent papers address these questions, focusing on TDP-43’s stability and protein-protein interactions. In PNAS online this week, researchers report that TDP-43 mutants last longer in the cell than the wild type protein, and are more likely to hook up with FUS, another protein involved in ALS. The second study, posted online in the Journal of Biological Chemistry on June 16, suggests one possible mechanism by which TDP-43 might bind other proteins. The authors report that TDP-43 possesses a prion-like domain that allows it to bind polyglutamate inclusions, such as those found in Huntington disease.
For the work described in PNAS, first author Shuo-Chien Ling, in the laboratory of Don Cleveland at the University of California in San Diego, led the effort to distinguish the actions of wild-type and mutant TDP-43. Ling examined three disease-associated TDP-43 mutants—G298S, Q331K, and M337V—as well as wild-type constructs. He used stably transfected HeLa cells to express a single copy of each transgene, in the same locus in each case, controlle by the inducible Tet operator and drive by a cytomegalovirus promoter.
TDP-43 pathology includes cytoplasmic aggregates of the normally nuclear protein. Ling and colleagues wondered if the mutant proteins hang around in the cell longer than the wild-type does. They pulse-labeled growing cells with radioactive methionine and cysteine to tag newly translated proteins, then followed TDP-43 radioactivity over time. The mutants lasted two to four times longer than wild-type TDP-43 does, suggesting they are more resistant to degradation. To test if this is true in humans, Ling and colleagues performed a similar experiment with primary fibroblasts. In wild-type fibroblasts, TDP-43 has a four-hour half-life, but this rises to 11 hours in cells from a person carrying the G298S mutation.
A longer half-life could dramatically affect TDP-43 biology, which is poorly characterized. To find clues to the protein’s normal role, the scientists isolated the wild-type protein, with its associated hangers-on, and used mass spectrometry to identify the partners. TDP-43 is known to modify mRNAs, so it was no surprise to see components of the heterogenous nuclear ribonuclear (hnRNP) protein complex, as well as other RNA-binding proteins. However, the researchers did not expect the second set of proteins attached to TDP-43: elements of the Drosha microprocessing complex that refines microRNA (miRNA). Perhaps, they write, TDP-43 may help bridge the two (hnRNP and miRNA) complexes, though it is not yet clear what function this might have. “I think it is a very tempting idea that it [TDP-43] is sitting on both pathways…to regulate gene expression,” Ling said.
Another interesting partner—FUS—appeared in the mass spec analysis. Researchers have long sought to connect TDP-43 and FUS in a common pathway leading to neurodegenerative disease, so Ling and colleagues pursued this particular interaction further. In co-immunoprecipitation experiments, FUS and TDP-43 did come together, but only weakly. Precipitating TDP-43 pulled down less than 1 percent of the cell’s FUS, and precipitating FUS similarly brought down less than 1 percent of TDP-43 protein. However, when the researchers performed the same experiments with the TDP-43 mutants Q331K and M337V, they found that much more FUS tagged along, suggesting the mutants have increased affinity for FUS, or a for complex that also contains FUS.
“The interaction is most likely an early event” in disease, Ling speculated, leading up to later stages where TDP-43 and FUS, normally nuclear proteins, are mislocalized and aggregated in the cytoplasm. Theoretically, he suggested, the same might happen in sporadic ALS, if some other mechanism enhanced the stability of wild-type TDP-43. Ling plans to examine TDP-43 stability and protein-protein interactions in neurons and patient-derived cells.
This paper is one of the first to show real differences between wild-type and mutant TDP-43, said Robert Baloh, Washington University School of Medicine in St. Louis, Missouri. The findings might also help explain why rodent models expressing wild-type human TDP-43 get sick, he said (see ARF related news story on Wils et al., 2010 and ARF related news story on Tatom et al., 2009). If there is excess TDP-43 production in those models, it might mimic the effects of abnormally stable TDP-43.
In addition, Ling and colleagues’ paper describes one possible way TDP-43 and FUS pathways could intersect. Both have prion-like, protein-protein interaction domains, he noted, by which they could potentially join the same complex.
Baloh, along with first author Rodrigo Fuentealba, led an inquiry into the prion-like region of TDP-43, which they describe in the recent Journal of Biological Chemistry paper. “We were just trying to find out what caused TDP-43 to leave the nucleus,” Baloh said. They theorized that TDP-43’s cytosolic presence might be due to the stress caused by other misfolded proteins accumulating there. Accordingly, they expressed a collection of aggregation-prone proteins in HeLa cells. Some, such as mutants of dynactin and calveolin, did not recruit TDP-43, but a polyglutamine protein-cyan fluorescent protein chimera (Q80-CFP) did. Similarly, a huntingtin fragment with an expanded polyglutamine region (Htt-Q72) attracted TDP-43. “It is very specific,” Baloh said. “TDP-43 gets sucked into the polyglutamine aggregates, but not into other types of aggregates.”
The researchers wondered which part of TDP-43 might bind to polyglutamine constructs, and engineered deletion constructs to find out. The carboxyl-terminal domain of TDP-43—where most ALS-causing mutations reside—was crucial for the interaction with Q80-CFP. In particular, Baloh and colleagues discovered that a region between amino acids 320 and 367 was needed to aggregate with polyglutamine.
Among proteins that bind polyglutamine, combined glutamine/asparagine regions are common. In TDP-43, the area that the scientists defined as polyglutamine-binding was 31 percent glutamine and asparagine. Prions—contagious, agglutinating proteins—have similar domains, Baloh noted. “That does not mean that TDP-43 is a prion,” he was quick to add. However, he said, their research suggests that TDP-43, when sequestered in polyglutamine aggregates, might be prevented from doing its job in the nucleus.
If so, the authors speculated, then TDP-43 overexpression might relieve the toxicity caused by polyglutamine proteins such as huntingtin. Experimenting with COS7 cells, they found that expression of Htt-Q72, tagged with the red fluorescent Cherry protein, killed half of the cell population within three days. When the researchers co-expressed excess TDP-43 along with the Htt-Q72-Cherry, survival rates went up, to approximately 70 percent. TDP-43 overexpression brought the cells up to nearly the same survival rate as control cells carrying only the Cherry protein—80 percent. Therefore, the authors suggest that TDP-43 overexpression partially rescues cells from huntingtin’s damaging effects, perhaps because there it provides sufficient TDP43 to function in the nucleus.
Other scientists have suggested that many of neurodegeneration’s usual suspects—β amyloid, tau, and α-synuclein, for example—may function in a prion-like or “prionoid” manner, spreading their toxic effects across the brain by converting normal proteins (reviewed in Cushman et al., 2010 and ARF Live Discussion). TDP-43 does not form amyloid structures, Baloh said, but it is still possible it has an alternative, “infectious” conformation. If so, the conversion of TDP-43 might explain why ALS frequently begins in one part of the body and spreads to the rest in a linear fashion. FUS, too, has a glutamate/asparagine-heavy region and co-aggregates with huntingtin (Doi et al., 2008) in perhaps another link between the two ALS proteins.
Although Baloh’s group experimented with cultured, non-neuronal cell lines, he pointed out that TDP-43 and huntingtin proteins associate in people with Huntington disease as well. In 2008, researchers at the University of British Columbia in Vancouver reported that TDP-43 is part of some cytoplasmic huntingtin inclusions in autopsy brain tissue samples (Schwab et al., 2008). However, the senior author of that study, Patrick McGeer, questioned the relevance of Baloh’s findings to human disease. There is no evidence, he said, that TDP-43’s mutations or its glutamate/asparagine region have any role in Huntington disease. Even if that is the case, Baloh wrote in an e-mail to ARF, the prion-like region of TDP-43 may be important for its function in both health and disease.—Amber Dance.
Ling SC, Albuquerque CP, Han JS, Lagier-Tourenne C, Tokunaga S, Zhou H, Cleveland DW. ALS-associated mutations in TDP-43 increase its stability and promote TDP-43 complexes with FUS/TLS. Proc Natl Acad Sci U S A. 2010 Jul 12. Abstract
Fuentalba RA, Udan M, Bell S, Wegorzewska I, Shao J, Diamond MI, Weihl CC, Baloh RH. Interaction with polyglutamine aggregates reveals a Q/N rich domain in TDP-43. J Biol Chem. 2010 Jun 16. Abstract