Masking a Splicing Site Compensates for TDP-43 Dysfunction

In TDP-43 proteinopathies, the RNA/DNA binding protein leaves its post in the nucleus and travels to cytosol, where it’s prone to aggregate. This meandering allows mis-spliced transcripts to accumulate. Could mis-splicing be prevented without restoring TDP-43 to the nucleus? Yes, say researchers led by Don Cleveland and Ze’ev Melamed at the University of California, San Diego, and Clotilde Lagier-Tourenne of Harvard Medical School, Boston—at least for one protein, stathmin-2. In the March 16 Science, they reported that in human motor neurons, antisense oligonucleotides slashed production of mis-spliced STMN2, boosted the full-length protein, and enabled axon regeneration. When injected into the cerebrospinal fluid of mice that mis-splice STMN2, the ASOs restored stathmin-2 levels.

  • TDP-43 prevents mis-splicing of STMN2 by blocking a cryptic-exon splice site.
  • Antisense oligonucleotides that bind near the site do the same.
  • The ASOs restored full-length STMN2 in TDP-43 mutant motor neurons.

“That cryptic mis-splicing can be targeted with ASOs is very exciting and promising from a therapeutic point of view,” Manuela Neumann, University of Tübingen, Germany, wrote (full comment below). “However, it remains to be seen if targeting only one of the many mis-spliced RNA targets caused by loss of nuclear TDP-43 will be sufficient to modulate the clinical disease course of TDP-43 proteinopathies.”

A microtubule-binding protein, stathmin-2 ensures neuron growth and survival (Jan 2019 news). Without it, neurons cannot extend or repair axons, nor do their lysosomes traverse axons properly (see graphic below). In ALS, loss of nuclear TDP-43 leads to a mess of mis-spliced transcripts. Like most, mis-spliced STMN2 is destroyed, reducing the amount of functional stathmin-2, though scientists recently found that some “cryptic transcripts” persist in TDP-43 linked ALS/FTD, yielding novel dysfunctional proteins, which can be detected in the CSF (Feb 2023 news).

In Sickness and in Health. Compared to healthy neurons (left), neurons lacking nuclear TDP-43 (right) do not make functional stathmin-2, leading to poor lysosomal trafficking and stunted axon regeneration. [Courtesy of O'Brien and Mizielinska, Science, 2023.]

Cleveland and colleagues previously reported that TDP-43 binds to a 24-residue stretch in exon 2a of STMN2 containing three GUGUGU repeats. This sits right next to the cryptic splice initiation site (see image below; Mar 2011 news). To find out how this motif controls splicing, co-first authors Melamed and Michael Baughn at UCSD replaced it with a sequence that binds the bacteriophage coat protein MS2. This protein plays no role in splicing, but it allowed the researchers to test if simply crowding the cryptic splice site shields it from the splicing machinery. They tested the strategy in human SH-SY5Y neuroblastoma cells.

As expected, cells carrying one MS2-STMN2 hybrid allele lacking the TDP-43 binding motif made cryptic transcripts and only half as much full-length STMN2 mRNA as control neurons. Surprisingly, after the scientists transfected the cells with viruses carrying MS2 genes, cryptic STMN2 transcripts plummeted by 90 percent, and full-length STMN2 levels were restored. “It is somewhat surprising that the TDP43 protein is not required for proper STMN2 splicing. As long as something … is sitting on the same locus, splicing of STMN2 exon 2A will be repressed,” noted Sami Barmada, University of Michigan, Ann Arbor. “This is incredibly important for the design of therapeutic modalities that act on this locus,” he added (comment below).

Could antisense oligonucleotides do the job? Interest in using ASOs for neurodegenerative diseases grew after the FDA approved nusinersen to treat spinal muscular atrophy in 2016. This ASO corrects mis-splicing of the SMN2 gene, increasing the motor neuron protein in the spinal cord (Finkel et al., 2016; Nov 2016 news).

STMN2 ASOs. Within exon 2a of the STMN2 gene, TDP-43 binds between a cryptic splice initiation site and a cryptic poly-A motif. The five ASOs that best squelched cryptic SMN2 splicing bind around the TDP-43 site. [Courtesy of Baughn et al., Science, 2023.]

Baughn and Melamed created 18-nucleotide-long ASOs spanning the entire cryptic exon. They tested 250 of them in human neuroblastoma cells that carried two copies of the ALS/FTD-linked N352S TDP-43 mutation. This single amino acid substitution causes partial loss of TDP43 function, resulting in cryptic STMN2 splicing. Of the 250 ASOs, the authors settled on five that best reduced cryptic STMN2 and restored full-length protein (see image above).

In human induced motor neurons that had TDP-43 knocked down, the best ASO, called rASO-5, increased full-length STMN2 mRNA from 30 to 86 percent of normal levels. It also normalized the amount of stathmin-2, despite the lack of nuclear TDP-43. Of the five ASOs tested, this oligonucleotide bound farthest from the TDP-43 binding motif. Barmada found that intriguing. “The most effective splice repressors were not those that closely overlap TDP43 binding sites. Rather, binding downstream of the TDP43 binding site produces the most impressive effects on STMN2 RNA and protein,” he noted

ASO treatment also helped human motor neurons rebound from injury. Lysosomes remobilized and axons that had been cut off regrew to wild-type length (see image below).

Dendritic Rescue. Compared to wild-type human induced motor neurons (left), TDP-43 knockdowns (middle) lacked stathmin-2 (green) and failed to regenerate axons (red) after they were severed (dashed white line). Neurons given an STMN2 ASO (right) expressed stathmin-2 and their axons regrew. [Courtesy of Baughn et al., Science, 2023.]

Would the ASOs work in vivo? To find out, the researchers created mice carrying one murine copy of STMN2 without a TDP-43 binding site. This caused chronic cryptic splicing, halving the amount of full-length STMN2 mRNA, though TDP-43 was unaffected and remained in the nucleus.

The scientists tested three of the ASOs by injecting them into the intracerebral ventricles of 2-month-old mice. Two weeks later, they tested SMNT2 splicing. Once again, rASO-5 performed best, suppressing cryptic STMN2 splicing by about half in the cortex and one-third in the spinal cord. It bumped up full-length mRNA and stathmin-2 protein levels to 75 percent of those in wild-type mice.

Could rASO5, or another of these ASOs, become a therapy for ALS and FTD? Researchers are cautious. Barmada noted that the mice that mis-spliced STMN2 had no apparent symptoms to monitor, making it difficult to gauge any potential therapeutic effect. Philip Wong of Johns Hopkins University in Baltimore wondered whether restoring repression of STMN2 cryptic exons alone would be enough to attenuate neurodegeneration. STMN2 knockout mice exhibit some, but not all, features of motor neuron disease, hinting that other cryptic exons contribute to disease pathology and symptoms (comment below).

Lagier-Tourenne thinks that the field needs mouse models that recapitulate TDP-43 mislocalization and motor symptoms to better study potential drugs. In the meantime, she and Cleveland are partnering with Ionis Pharmaceuticals to continue screening these STMN2 ASOs for clinical use.

Another ASO for STMN2 is already being tested in ALS. The Cambridge, Massachusetts-based QurAlis, co-founded by Kevin Eggan at Harvard University, began a Phase 1 clinical trial to test the their lead compound, QRL-201 (company press release). It is not clear how the ASO works.—Chelsea Weidman Burke

Comments

  1. Cryptic exon mis-splicing due to loss of nuclear TDP-43 function has emerged as one of the hallmarks in TDP-43 proteinopathies. One previously identified target is STMN2 RNA, which is particularly interesting due to the crucial role of STMN2 in axonal regeneration. The study by Baughn et al. deciphers in detail the molecular mechanism through which TDP-43 regulates STMN2 pre-mRNA processing and identified antisense oligonucleotides (ASOs) capable of restoring normal stathmin-2 protein and RNA levels within the mammalian nervous system. I congratulate the authors on this exciting and very thoroughly performed study.

    Obviously, that cryptic mis-splicing, per se, can be targeted with ASOs is very exciting and promising from a therapeutic point of view. Although, as pointed out by the authors, it remains to be seen if targeting only one of the many mis-spliced RNA targets due to nuclear TDP-43 loss will be sufficient to modulate the clinical disease course of TDP-43 proteinopathies.

    View all comments by Manuela Neumann

  2. This is compelling work, for a number of reasons. First, it is somewhat surprising that the majority of the TDP-43 protein is not required for proper STMN2 splicing. As long as something (MS2 coat protein, CasRx, an ASO) is sitting on the same locus, splicing of STMN2 exon 2A will be repressed. This is incredibly important for the design of therapeutic modalities that act on this locus.

    Just as important, the authors found that restoration of STMN2 protein to 25 percent of normal is sufficient to rescue phenotypes such as axonal regeneration. This means that therapies may not have to completely replace STMN2, as effects may be seen with as little as a quarter of the normal STMN2 levels.

    The positional effects of ASOs are intriguing. The most effective splice repressors were not those that closely overlap TDP-43 binding sites. Rather, binding downstream of the TDP-43 binding site produces the most impressive effects on STMN2 RNA and protein.

    Although the authors demonstrated restoration of STMN2 levels in vivo (in humanized mice) through injection of rASOs, there were no clear symptoms to monitor in these animals. This makes it difficult to gauge the potential therapeutic impact of this strategy.

    Original studies hinted at STMN2 loss with TDP43 overexpression, in addition to loss of function. In this study, however, expression of a mutant (Q331K) TDP-43 that remains in the nucleus had little to no effect on STMN2 processing in STMN2 humanized mice. This result argues against STMN2 mis-splicing in TDP-43 transgenic and/or gain-of-function models.

    View all comments by Sami Barmada

  3. Baughn et al. address the molecular mechanism of splicing repression by TDP-43 by focusing on the inclusion of a non-conserved cryptic exon embedded within intron 1 of STMN2 (Klim et al., 2019; Melamed et al., 2019) and the identification of ASOs to restore its repression as a potential therapeutic strategy for ALS. Using a clever gene-editing approach, the authors showed that TDP-43 binding to UG-rich repeats within this cryptic exon blocks recognition of the 3' splice site in intron 1 to allow proper splicing of STMN2 pre-mRNA.

    While this is an attractive mechanism of splicing repression by TDP-43, it is not clear how ALS-linked missense mutant TDP-43 (occurs mostly within C-terminal prion-like domain of TDP-43) apparently fails to repress splicing of this STMN2 cryptic exon (Klim et al., 2019; Melamed et al., 2019) How would one amino acid substitution within the C-terminal domain of TDP-43 release this steric blockage of the recognition of the 3' splice site in intron 1 of STMN2? One resolution is if there exists an instructive repression signal directed to the spliceosome conferred by the C-terminal domain of TDP-43; it is possible that such a signal can be disrupted by missense mutations within this domain. Structural/functional analysis may be necessary to resolve this issue.

    Because TDP-43-dependent cryptic exons, including that of STMN2, are non-conserved (Ling et al., 2015), Baughn et al. partially humanized the mouse Stmn2 to validate a therapeutic ASO designed to restore repression of STMN2 cryptic exon. While their data support such ASOs for clinical testing, which recently has already begun (see clinicaltrials.gov), an outstanding question is whether restoring STMN2 cryptic exon repression is sufficient to attenuate neuron loss within the context of neurodegeneration. That genetic knockout/knockdown studies of Stmn2 in mice, which exhibit some, but not all, key features of motor neuron disease, namely motor neuron loss leading to denervation muscle atrophy and paralysis (Guerra San Juan et al., 2022; Lopez-Erauskin et al., 2022), suggests that targeting other TDP-43 dependent cryptic exons, such as UNC13A (Ma et al., 2022; Brown et al., 2022) and other emerging sites, would be required.

    If multiple TDP-43-dependent cryptic exons are required to be repressed, the N-terminal half of TDP-43 (containing its RNA binding domains—RRM1 and RRM2) fused to the repressor domain of RAVER1 (Ling et al., 2015; Donde et al., 2019) might prove to be beneficial. Since it is now recognized that loss of TDP-43 splicing repression can occur during presymptomatic stage of ALS-FTD (Seddighi et al., 2023; Irwin et al., 2023), early intervention to complement such loss would be desirable. This unique repressor, shown to complement the loss of TDP-43 splicing repression, when delivered early by an adeno-associated viral vector, would hold promise for the treatment of ALS-FTD.

     

    References:

    Klim JR, Williams LA, Limone F, Guerra San Juan I, Davis-Dusenbery BN, Mordes DA, Burberry A, Steinbaugh MJ, Gamage KK, Kirchner R, Moccia R, Cassel SH, Chen K, Wainger BJ, Woolf CJ, Eggan K. ALS-implicated protein TDP-43 sustains levels of STMN2, a mediator of motor neuron growth and repair. Nat Neurosci. 2019 Feb;22(2):167-179. Epub 2019 Jan 14 PubMed.

    Melamed Z, López-Erauskin J, Baughn MW, Zhang O, Drenner K, Sun Y, Freyermuth F, McMahon MA, Beccari MS, Artates JW, Ohkubo T, Rodriguez M, Lin N, Wu D, Bennett CF, Rigo F, Da Cruz S, Ravits J, Lagier-Tourenne C, Cleveland DW. Premature polyadenylation-mediated loss of stathmin-2 is a hallmark of TDP-43-dependent neurodegeneration. Nat Neurosci. 2019 Feb;22(2):180-190. Epub 2019 Jan 14 PubMed.

    Ling JP, Pletnikova O, Troncoso JC, Wong PC. NEURODEGENERATION. TDP-43 repression of nonconserved cryptic exons is compromised in ALS-FTD. Science. 2015 Aug 7;349(6248):650-5. PubMed.

    Guerra San Juan I, Nash LA, Smith KS, Leyton-Jaimes MF, Qian M, Klim JR, Limone F, Dorr AB, Couto A, Pintacuda G, Joseph BJ, Whisenant DE, Noble C, Melnik V, Potter D, Holmes A, Burberry A, Verhage M, Eggan K. Loss of mouse Stmn2 function causes motor neuropathy. Neuron. 2022 May 18;110(10):1671-1688.e6. Epub 2022 Mar 15 PubMed.

    Lopez-Erauskin J, Bravo-Hernandez M, Presa M, Baughn MW, Melamed Z, Beccari MS, AgradeAlmeidaQuadros AR, Zuberi A, Ling K, Platoshyn O, Nino-Jara E, Ndayambaje IS, Arnold-Garcia O, McAlonis-Downes M, Cabrera L, Artates JW, Ryan J, Bennett F, Jafar-nejad P, Rigo F, Marsala M, Lutz CM, Cleveland DW, Lagier-Tourenne C. Stathmin-2 loss leads to neurofilament-dependent axonal collapse driving motor and sensory denervation. 2022 Dec 12 10.1101/2022.12.11.519794 (version 1) bioRxiv.

    Ma XR, Prudencio M, Koike Y, Vatsavayai SC, Kim G, Harbinski F, Briner A, Rodriguez CM, Guo C, Akiyama T, Schmidt HB, Cummings BB, Wyatt DW, Kurylo K, Miller G, Mekhoubad S, Sallee N, Mekonnen G, Ganser L, Rubien JD, Jansen-West K, Cook CN, Pickles S, Oskarsson B, Graff-Radford NR, Boeve BF, Knopman DS, Petersen RC, Dickson DW, Shorter J, Myong S, Green EM, Seeley WW, Petrucelli L, Gitler AD. TDP-43 represses cryptic exon inclusion in the FTD-ALS gene UNC13A. Nature. 2022 Mar;603(7899):124-130. Epub 2022 Feb 23 PubMed.

    Brown AL, Wilkins OG, Keuss MJ, Hill SE, Zanovello M, Lee WC, Bampton A, Lee FC, Masino L, Qi YA, Bryce-Smith S, Gatt A, Hallegger M, Fagegaltier D, Phatnani H, NYGC ALS Consortium, Newcombe J, Gustavsson EK, Seddighi S, Reyes JF, Coon SL, Ramos D, Schiavo G, Fisher EM, Raj T, Secrier M, Lashley T, Ule J, Buratti E, Humphrey J, Ward ME, Fratta P. TDP-43 loss and ALS-risk SNPs drive mis-splicing and depletion of UNC13A. Nature. 2022 Mar;603(7899):131-137. Epub 2022 Feb 23 PubMed. Correction.

    Donde A, Sun M, Ling JP, Braunstein KE, Pang B, Wen X, Cheng X, Chen L, Wong PC. Splicing repression is a major function of TDP-43 in motor neurons. Acta Neuropathol. 2019 Nov;138(5):813-826. Epub 2019 Jul 22 PubMed.

    Seddighi S, Qi YA, Brown AL, Wilkins OG, Bereda C, Belair C, Zhang Y, Prudencio M, Keuss MJ, Khandeshi A, Pickles S, Hill SE, Hawrot J, Ramos DM, Yuan H, Roberts J, Sacramento EK, Shah SI, Nalls MA, Colon-Mercado J, Reyes JF, Ryan VH, Nelson MP, Cook C, Li Z, Screven L, Kwan JY, Shantaraman A, Ping L, Koike Y, Oskarsson B, Staff N, Duong DM, Ahmed A, Secrier M, Ule J, Jacobson S, Rohrer J, Malaspina A, Glass JD, Ori A, Seyfried NT, Maragkakis M, Petrucelli L, Fratta P, Ward ME. Mis-spliced transcripts generate de novo proteins in TDP-43-related ALS/FTD. bioRxiv. January 23, 2023 bioRxiv

    Irwin KE, Jasin P, Braunstein KE, Sinha I, Bowden KD, Moghekar A, Oh ES, Raitcheva D, Bartlett D, Berry JD, Traynor B, Ling JP, Wong PC. A fluid biomarker reveals loss of TDP-43 splicing repression in pre-symptomatic ALS. bioRxiv. January 24, 2023 bioRxiv

    View all comments by Philip Wong

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Microtubule Regulator Connects TDP-43 to Axonal Dysfunction
  2. Can ‘Cryptic Peptides’ Peg People with TDP-43 Pathology?
  3. CLIPs of TDP-43 Provide a Glimpse Into Pathology
  4. Positive Trials of Spinal Muscular Atrophy Bode Well for Antisense Approach

Paper Citations

  1. Finkel RS, Chiriboga CA, Vajsar J, Day JW, Montes J, De Vivo DC, Yamashita M, Rigo F, Hung G, Schneider E, Norris DA, Xia S, Bennett CF, Bishop KM. Treatment of infantile-onset spinal muscular atrophy with nusinersen: a phase 2, open-label, dose-escalation study. Lancet. 2016 Dec 17;388(10063):3017-3026. Epub 2016 Dec 7 PubMed.

External Citations

  1. Phase 1
  2. company press release

Further Reading

Papers

  1. Lopez-Erauskin J, Bravo-Hernandez M, Presa M, Baughn MW, Melamed Z, Beccari MS, AgradeAlmeidaQuadros AR, Zuberi A, Ling K, Platoshyn O, Nino-Jara E, Ndayambaje IS, Arnold-Garcia O, McAlonis-Downes M, Cabrera L, Artates JW, Ryan J, Bennett F, Jafar-nejad P, Rigo F, Marsala M, Lutz CM, Cleveland DW, Lagier-Tourenne C. Stathmin-2 loss leads to neurofilament-dependent axonal collapse driving motor and sensory denervation. 2022 Dec 12 10.1101/2022.12.11.519794 (version 1) bioRxiv.

Primary Papers

  1. Baughn MW, Melamed Z, López-Erauskin J, Beccari MS, Ling K, Zuberi A, Presa M, Gonzalo-Gil E, Maimon R, Vazquez-Sanchez S, Chaturvedi S, Bravo-Hernández M, Taupin V, Moore S, Artates JW, Acks E, Ndayambaje IS, Agra de Almeida Quadros AR, Jafar-Nejad P, Rigo F, Bennett CF, Lutz C, Lagier-Tourenne C, Cleveland DW. Mechanism of STMN2 cryptic splice-polyadenylation and its correction for TDP-43 proteinopathies. Science. 2023 Mar 17;379(6637):1140-1149. Epub 2023 Mar 16 PubMed.
  2. O'Brien N, Mizielinska S. A cryptic clue to neurodegeneration?. Science. 2023 Mar 17;379(6637):1090-1091. Epub 2023 Mar 16 PubMed.

Annotate

To make an annotation you must Login or Register.