29 Feb 2024

TDP-43 leads a varied life, interacting with a host of different RNAs and proteins as it shuttles from the nucleus to the cytoplasm and back. This offers this RNA-binding protein plenty of opportunities to take up with the wrong crowd. Case in point: recent papers identify two binding partners that maroon TDP-43 in the cytoplasm, where it aggregates. One, 14-3-3θ, is a scaffolding protein that, under physiological conditions, may stabilize TDP-43 during its travels to the cytoplasm. However, if TDP-43 is mutated, or if 14-3-3θ levels run amok, their relationship turns toxic. Led by Lars Ittner of Macquarie University in Sydney and published in Neuron on February 6, the study casts 14-3-3θ as a potential therapeutic target.

The second partner, reported online in Nature on November 8, 2023, is methylated CAG repeat RNA—the kind that underlies triplet repeat diseases such as Huntington’s and certain ataxias. Scientists led by Yinsheng Wang at the University of California, Riverside, found that this methylated RNA entangled TDP-43 in stress granules. Together, the two studies add to the mounting list of mechanisms at play in TDP-43-mediated neurodegeneration, albeit in different disease contexts, and hint at fresh ways to intervene.

TDP-43 plays an indispensable role in the splicing and proper transport of an ever-growing list of transcripts. Cytoplasmic aggregates of the protein were first identified in FTD and ALS. However, they also turn up in Alzheimer’s, Parkinson’s, and Huntington’s, and form the basis of limbic predominant TDP-43 encephalopathy (LATE). What forces steer TDP-43 to aggregate in the first place?

Co-first authors Yazi Ke and Annika van Hummel and colleagues in Ittner’s group approached this question by screening for proteins that bind TDP-43’s glycine-rich domain, which houses the majority of its known pathogenic mutations. The top hit was 14-3-3θ, a member of the 14-3-3 family of scaffolding proteins that take part in all manner of signaling and protein trafficking.

The researchers investigated the relationship in different models and disease scenarios. In short, they identified a 10-amino-acid stretch within 14-3-3θ’s sixth α-helix that bound to TDP-43. Nixing this binding site led to rapid distribution of TDP-43 to the cytoplasm once cells were put under stress, suggesting that the interaction somehow facilitates TDP-43’s return to the nucleus. In keeping with this, the interaction diminished when TDP-43 bound to RNA.

Curiously, pathogenic mutations in TDP-43 strengthened the interaction, suggesting that too much TDP-43/14-3-3θ binding is as bad as too little. Indeed, overexpressing 14-3-3θ in mice proved detrimental. TDP-43 accumulated in the cytosol in spinal cord neurons (image below), and by 12 months their grip weakened and they struggled to hang upside down on a wire mesh. In the hippocampi of TDP-43 A315T mice, 14-3-3θ overexpression ramped up cytoplasmic accumulation of TDP-43 in neurons, half of which died.

Stuck in the Cytoplasm. In wild-type mice, TDP-43 (red) shifted from nucleus into cytoplasm in spinal cord motor neurons transduced with extra 14-3-3θ (green, open arrows). In non-transduced cells, TDP-43 stayed in the nucleus (filled arrows). [Courtesy of Ke et al., Neuron, 2024.]

A now-old study had found excess 14-3-3θ transcripts in brain samples from people with FTD/ALS (Malaspina et al., 2000). Indeed, Ke found a glut of 14-3-3θ protein in postmortem brain from people with sporadic FTLD, as well as in motor neurons from people with sporadic ALS. In both cases, 14-3-3θ tagged along with TDP-43 in the cytoplasm of neurons. In people who had had a neurodegenerative disease but no TDP-43 pathology, including FTLD-tau, AD, and ALS cases caused by SOD1 mutation, 14-3-3θ levels were near those of healthy controls.

Curiously, 14-3-3θ levels were normal in people who had had familial ALS caused by hexanucleotide expansions in the C9orf72 gene, despite the presence of TDP-43 pathology in those cases. In these cases, RNA foci and poly dipeptide repeats might drive TDP-43 pathology, the authors suggested.

 

Partners in Crime. Proximity ligation assay, which lights up when 14-3-3θ and TDP-43 meet, catches both proteins co-mingling in the cytoplasm of cells in a brain sample from a person with FTLD-TDP (left), but not in a control (right).

Finally, the authors used 14-3-3θ’s binding to TDP-43 to try a treatment strategy. Essentially, they dangled the requisite α-helix as bait to capture and destroy TDP-43 in the cytoplasm. They affixed an FK506-binding protein degradation domain to this 14-3-3θ helix, then expressed it in neurons and various mouse models (for a review of these degradation domains, see An et al. 2015). In all these models, the degradation domain functioned—reducing TDP-43 pathology, sparing neurons, and preventing cognitive and motor deficits.

Attracted to Methylation
In their study, Wang and colleagues came across the other toxic TDP-43 liaison. Co-first authors Yuxiang Sun and Hui Dai did not set out to look for TDP-43 interactors, but instead were investigating CAG repeat expansions. Both repeat-laden transcripts that aggregate and sequester RNA-binding proteins, and polyglutamine proteins translated from these transcripts, wreak havoc on cells. Focusing on the former, Sun found that their adenosines could be methylated, and that the longer the repeat, the more methyl groups piled on.

 

The researchers pegged TRMT61A as the methyltransferase responsible for these additions, and ALKBH3 as the demethylase that removes them. What’s more, they found that longCAG repeats doused expression of ALKBH3, explaining the link between repeat length and CAG methylation. Notably, overexpressing the demethylase in worms and flies prevented neurodegeneration caused by the CAG repeats and increase lifespan (image at left).

What does this have to do with TDP-43? TDP-43 bound to CAG repeats, and methylated adenosine (m1A) boosted this interaction. In cell culture, fly, and worm models, the researchers showed that m1A-CAG repeat RNA made the protein phase-separate into gel-like droplets, from where it accumulated in cytoplasmic stress granules. This type of phase transition was reported previously for RNA-binding and polyglutamine proteins, including TDP-43 (April 2018 news; Mar 2019 news). The authors believe this m1A-CAG RNA/TDP-43 connection might occur in addition to problems caused by polyglutamine peptides translated from the RNA repeats.—Jessica Shugart

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Liquid Phase Transition: A Deluge of Data Points to Multiple Regulators 21 Apr 2018

Paper Citations

1. Malaspina A, Kaushik N, de Belleroche J. A 14-3-3 mRNA is up-regulated in amyotrophic lateral sclerosis spinal cord. J Neurochem. 2000 Dec;75(6):2511-20. PubMed.
2. An W, Jackson RE, Hunter P, Gögel S, van Diepen M, Liu K, Meyer MP, Eickholt BJ. Engineering FKBP-Based Destabilizing Domains to Build Sophisticated Protein Regulation Systems. PLoS One. 2015;10(12):e0145783. Epub 2015 Dec 30 PubMed.

Other Citations

1. Mar 2019 news

Further Reading

No Available Further Reading