For toxicity, “You need cleavage, coupled with another downstream toxic insult,” said first author Scott Pesiridis, who worked with senior author Virginia Lee at the University of Pennsylvania in Philadelphia. He found that messing with cellular transport, destroying RNAs, or seeding aggregates provided the crucial second hit leading to CTF inclusions. Corroborating these results is another paper posted by the journal. Researchers from Japan report that seeding with a few CTF fibrils recruited TDP-43 into aggregates both in vitro and in vivo. First author Yoshiaku Furukawa of Keio University led the work along with senior author Nobuyuki Nukina at the Riken Brain Science Institute in Wako.

Since TDP-43’s natural protease is unknown, Pesiridis used his own: the Tobacco Etch Virus (TEV) protease. He cloned the TEV cleavage site into several locations and found that sites within the RRM2 RNA-binding domain were inaccessible to the protease. He also discovered that the RRM2 was essential for CTFs to remain in the cell long enough for him to observe them, so for further studies he used a TDP-43 gene that TEV split at amino acid 182, upstream of the RRM2. Furukawa and Nukina also found evidence for a protease-resistant area including the RRM2 domain.

Pesiridis’ CTF travelled to the cytoplasm, as expected, but it failed to replicate TDP-43 proteinopathy pathology. “The fragment was, surprisingly, not forming aggregates, so that was a big head-scratcher,” he recalled. Aaron Gitler, also of the University of Pennsylvania, but not involved in this study, noted that his and other labs have observed CTF aggregates when they express a fragmented TDP-43 gene (see ARF related news story on Johnson et al., 2008). The difference, Pesiridis suggested, might be that his CTFs have a normal development—they start as full-length protein, have a chance to mature, then are cleaved and expelled from the nucleus. Expressing CTFs alone skips those steps. “It seems like this is a more natural model of what happens to TDP-43,” agreed Gitler. Furukawa and colleagues took a different tack by transducing TDP-43 fibrils, which they made in vitro, into their human embryonic kidney cells. Only with these exogenous seeds present did endogenous TDP-43 begin to aggregate.

Pesiridis hypothesized that cellular stress or perturbations could precipitate aggregation, and tested three possible treatments: seeding, RNase treatment, and dynein inhibition to block protein transport. Like Furukawa, he found that if cells already contained cytoplasmic CTF seeds, the CTFs coming out of the nucleus would join in. This is reminiscent of how prions are transmitted, said Jim Shorter, another University of Pennsylvania researcher who did not participate in the study. TDP-43’s C-terminal domain, he noted, is similar to sequences found in yeast prions (Cushman et al., 2010).

Pesiridis suspects that once TDP-43 is cleaved, the CTFs attach to RNAs and other RNA-binding proteins to hitch a ride out of the nucleus. Once in the cytoplasm, he thinks, TDP-43’s continued association with these partners keeps it in a stable, soluble conformation that the cell can destroy. To test this hypothesis, he treated the cells with RNase. Destroying RNAs caused diffuse, dissolved CTFs to cluster together in inclusion-like structures. The results suggest RNAs are necessary to maintain soluble CTFs, although Pesiridis concedes that whole-cell RNAse treatment could have other effects, too.

RNA-protein complexes do not halt as soon as they reach the other side of the nuclear membrane; they are trafficked to the proper cellular location (see ARF related news story on Buckley et al., 2011). Thus, Pesiridis suspected that interfering with microtubule transport of RNAs would force CTFs into an unnaturally high concentration, where they might aggregate. Indeed, cells treated with dynein inhibitors evinced more cytoplasmic CTFs than their untreated counterparts.

Other situations might also serve as the second hit that induces CTF aggregation, Pesiridis said, suggesting oxidative stress or the formation of stress granules as possibilities. In another study, blocking autophagy boosted CTF levels (see ARF related news story), and researchers have found that TDP-43 inclusion formation is associated with cellular stress (Liu-Yesucevitz et al., 2010).

The question remains as to exactly what aggregated CTFs do in the cell once they form—they could cause disease, or merely be a sideshow to the real pathology processes. Or, as Furukawa suggested, they might even be protective by sequestering pathogenic TDP-43 forms. “Addition of seeds might even be therapeutic,” he theorized in an e-mail to ARF, though that remains to be proven.—Amber Dance.

References:
Pesiridis GS, Tripathy K, Tanik S, Trojanowski JQ, Lee VM. A “two hit” hypothesis for inclusion formation by carboxy terminal fragments of TDP-43 linked to RNA depletion and impaired microtubule dependent transport. J Biol Chem. 2011 Mar 24. Abstract

Furukawa Y, Kaneko K, Watanabe S, Yamanaka K, Nukina N. A seeding reaction recapitulates intracellular formation of sarkosyl-insoluble TAR DNA binding protein-43 inclusions. J Biol Chem. 2011 Mar 24. Abstract