In amyotrophic lateral sclerosis and frontotemporal dementia, loss of the RNA-binding protein TDP-43 from the nucleus creates a surge of mis-spliced mRNAs in neurons. So far, only one of these errant transcripts, stathmin-2, has been tied to disease pathology. Now, two preprints uploaded to bioRXiv on April 4 detail another—UNC13A. Variants in this gene increase risk for ALS/FTD.

  • Neurons lacking nuclear TDP-43 mis-splice UNC13A, make less of the protein.
  • UNC13A risk variants incorporate cryptic exons.
  • This only occurs in brain and spinal cord tissue harboring TDP-43 deposits.

One paper was penned by researchers led by Pietro Fratta, University College London, and Michael Ward, National Institute of Neurological Disorders and Stroke, Bethesda, Maryland. The second is from the labs of Aaron Gitler, Stanford University, California, and Leonard Petrucelli, Mayo Clinic, Jacksonville, Florida. Both papers report that if TDP-43 binding to UNC13A falters, the transcript is mis-spliced and less Unc13A protein is made. In frontal cortex neurons from people who had ALS/FTD with TDP-43 deposits, mis-splicing was highest in cells carrying the UNC13A risk variants. The studies support the idea that sequestering TDP-43 in the cytosol may unleash aberrant splicing events that contribute to neurodegenerative disease pathology.

Huda Zoghbi, Baylor College of Medicine, Houston, praised the thorough experiments from both groups. “They’ve linked two phenomena in ALS—TDP-43 accumulation and risk variants—together with a biological explanation,” she told Alzforum.

In almost all people with ALS and half of those with FTD, TDP-43 gets trapped in the cytosol, where it forms deposits (Sep 2020 news). As TDP-43 protein levels fall in the nucleus, splicing errors that it would normally prevent begin to magnify (reviewed by Klim et al., 2021; Mar 2011 news). One such snafu is the “cryptic exon.” This happens when a piece of an intron mistakenly incorporates into the mRNA, which can lead to cell dysfunction and death. In ALS, cryptic exons can run amok, leading researchers to search for those that might contribute to disease pathology (Aug 2015 news). 

Previously, scientists found that depleting TDP-43 from the nucleus prompts a cryptic exon to form in microtubule-binding protein stathmin-2, cutting its protein levels in half (Dec 2018 conference news; Jan 2019 news). The STMN2 cryptic exon occurred only in motor neurons from people with ALS, who had less stathmin-2 protein in their spinal cord tissue. What other errant transcripts might exacerbate ALS/FTD pathology?

To find out, both groups searched for splicing changes caused by the absence of TDP-43. From the Ward and Fratta labs, co-first authors Anna-Leigh Brown, Oscar Wilkins, and Matthew Keuss, all at UCL, and Sarah Hill at NINDS, used CRISPR to stifle TDP-43 expression in human induced pluripotent stem cell (iPSC)-derived neurons, then sequenced the RNA. They found 179 cryptic exons.

Co-first authors Rosa Ma in Gitler’s lab and Mercedes Prudencio and Yuka Koike in Petrucelli’s lab took a different approach to assess splicing variation. They reanalyzed RNA sequencing data from frontal cortex neurons in postmortem brain samples taken from seven older adults who had had FTD or FTD/ALS (Liu et al., 2019). They found 65 differently spliced transcripts in nuclei lacking TDP-43 that they deemed worth of further study. 

Splicing Differences. In human iPSC-derived neurons, knocking down TDP-43 (right) upped the percentage of transcripts containing cryptic exons, including those in UNC13A and stathmin-2 (STMN2). Each circle denotes a splice junction. [Courtesy of Brown et al., bioRXiv, 2021.]

The researchers were interested in UNC13A because of its association with ALS and FTD (Mar 2018 news; Apr 2018 news). Named for the uncoordinated movements of C. elegans when the gene is mutated, UNC13A is a member of a family of proteins that regulate axon function and neurotransmitter release. In mice, deleting UNC13A causes motor neuron defects (Sept 2009 news).

New Cryptic Exon. Compared to UNC13A splicing in control iPSC-derived neurons (teal), the transcript in cells with TDP-43 knocked down (gold) harbored a cryptic exon. Single-nucleotide polymorphisms (SNPs) in the intron and cryptic exon region (CE) close to the TDP-43 binding site (green square) made splicing even worse. [Courtesy of Brown et al., bioRxiv, 2021.]

Lo and behold, both groups identified the same cryptic exon between exons 20 and 21 (see image above). Like the STMN2 cryptic exon, the UNC13A one contains a stop codon, which the authors believe causes nonsense mediated decay of the mRNA. After knocking down TDP-43 in SH-SY5Y cells and iPSC-derived neurons, the cryptic exon appeared, while normal Unc13A mRNA and protein waned.

What about in people? Brown and colleagues analyzed the same human RNA-Seq dataset as Ma and colleagues did, finding cryptic exons only in nuclei devoid of TDP-43. Within medial frontal lobe neurons from four people with FTD, Ma and colleagues likewise saw the UNC13A cryptic exon only in nuclei lacking TDP-43 (see image below). “UNC13A risk alleles might act as an Achilles’ heel—lurking under the surface, not causing problems until TDP-43 becomes dysfunctional,” Ma and colleagues wrote in their preprint.

Exit TDP-43, Enter Cryptic Exon. In medial frontal lobe neurons from people who had FTD with TDP-43 deposits (top), cryptic exons (red) are incorporated into UNC13A transcripts (arrowheads) in nuclei lacking TDP-43 (green). In control tissue (bottom), no cryptic exons are made (arrows). [Courtesy of Ma et al., bioRxiv, 2021].

Looking more broadly, both groups sequenced RNA from brain and spinal cord tissue extracts taken from people with ALS or FTD and controls in the New York Genome Center (NYGC) ALS Consortium. They found the cryptic exon only in people who had accumulated TDP-43 deposits. Brown and colleagues detected the transcript in 89 percent of people who had FTD and 38 percent of those who had ALS. Ma and colleagues saw it in almost half of FTD cases. They also found it in 63 frontal cortex tissue samples from 49 FTD cases or healthy controls from the Mayo Clinic Brain Bank. In the NYGC samples, cryptic exon expression mirrored TDP-43 deposits, appearing in the spinal cords and motor cortices of ALS cases and frontal and temporal cortices in FTD cases.

Where Risk SNPs Come In
In the UNC13A gene, two single-nucleotide polymorphisms (SNPs) are known that increase a person's risk of ALS/FTD: rs12973192 in the cryptic exon and rs12608932 in the nearby TDP-43 binding region. Both research groups found that these variants weakened TDP-43 binding to UNC13A. Brown and colleagues believe this decreased binding promoted cryptic splicing.

How do these alleles increase disease risk? Looking through whole-genome sequencing (WGS) data from the NYGC cohort, Brown and colleagues found that people who were homozygous for either SNP had more UNC13A cryptic exon transcripts than did heterozygotes or noncarriers. The amount of UNC13A transcripts with the cryptic exon rose with that of STMN2 cryptic exons, which they previously found to correlate with TDP-43 deposits (Prudencio et al., 2020; commentary by Glass 2020). Ma and colleagues found that people with FTD/ALS who carry the rs12973192 risk SNP expressed the cryptic exon in almost every UNC13A transcript, while those with the common allele had up to three times less.

Understanding how loss of the UNC13A protein affects ALS/FTD will be the next step. “How much is this gene actually contributing to the demise of motor neurons?” Zoghbi asked. Ma and colleagues found evidence that the UNC13A variants accelerate disease progression. Among 205 people from the Mayo Clinic bank who had FTD with TDP-43 deposits, those who were homozygous for a risk haplotype that includes both UNC13A risk variants died sooner than those with no risk variant.

Mis-spliced STMN2 and UNC13A might be just the tip of the iceberg. “I am certain there will be others,” Zoghbi said. “The data imply it [cryptic exons] could be relevant to any disorder with loss of nuclear TDP-43 function,” she added. Philip Wong, Johns Hopkins University, Baltimore, agreed. “Other cryptic exons could also be influenced by SNPs in introns that modify TDP-43 binding,” he wrote to Alzforum. In Wong's group, the hunt for them is on (full comment below).—Chelsea Weidman Burke 


  1. Recent studies from human neurodegenerative disease strongly support the notion that nuclear depletion of TDP-43, and the resulting compromised splicing repression, represents an early event that contributes to disease pathogenesis. First, we documented loss of splicing repression in brains of ALS-FTD cases (Ling et al., 2015). Second, TDP-43 nuclear depletion in brain neurons has been reported at the presymptomatic stage in a C9ORF72-linked FTD-ALS patient, suggesting that loss of splicing repression represents an early event in disease progression (Vatsavayai et al., 2016). Third, incorporation of nonconserved cryptic exons can also be found in cases of Alzheimer’s disease with TDP-43 pathology in the absence of cytoplasmic inclusions (Sun et al., 2017). Fourth, ALS-linked mutant forms of TDP-43 fail to repress nonconserved cryptic exons (e.g., Stathmin-2), independent of TDP-43 cytoplasmic aggregation (Klim et al., 2019; Melamed et al., 2019). 

    Identification of UNC13A as a TDP-43 cryptic exon target that is linked to risk SNPs for ALS-FTD in these two manuscripts provides further support for the idea that loss of TDP-43 splicing repression of cryptic exons drives disease or influence its progression. 

    While depletion of TDP-43 would lead to inclusion of a host of cryptic exons, it is also possible that alterations in TDP-43 binding sites (UG rich element) could potentially impact on cryptic exon inclusion, as demonstrated by these two studies, for the SNP within a cryptic exon of UNC13A. As a “guardian of the transcriptome,” TDP-43, represses nonconserved cryptic exons, many of which reside within introns. This raises the important question as to whether other cryptic exon targets are also subject to influence of risk SNPs for ALS-FTD.

    Based on first principals, it is conceivable that a series of other cryptic exons could be influenced by SNPs, which often reside within introns, that modify the binding of TDP-43. Positive outcomes from future studies should validate this notion. 


    . NEURODEGENERATION. TDP-43 repression of nonconserved cryptic exons is compromised in ALS-FTD. Science. 2015 Aug 7;349(6248):650-5. PubMed.

    . Timing and significance of pathological features in C9orf72 expansion-associated frontotemporal dementia. Brain. 2016 Dec;139(Pt 12):3202-3216. Epub 2016 Oct 22 PubMed.

    . Cryptic exon incorporation occurs in Alzheimer's brain lacking TDP-43 inclusion but exhibiting nuclear clearance of TDP-43. Acta Neuropathol. 2017 Jun;133(6):923-931. Epub 2017 Mar 22 PubMed.

    . 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.

    . 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.

Make a Comment

To make a comment you must login or register.


News Citations

  1. In ALS and FTD, Two Different Routes to TDP-43 Aggregation
  2. CLIPs of TDP-43 Provide a Glimpse Into Pathology
  3. Does New Role for ALS-Linked Protein Help Explain Neurodegeneration?
  4. Beyond the Nucleus: TDP-43 Sticks Together, For Better or Worse
  5. Microtubule Regulator Connects TDP-43 to Axonal Dysfunction
  6. Massive ALS GWAS Cements Cytoskeletal Link to Disease
  7. Genetics Tie ALS into the Frontotemporal Dementia Spectrum
  8. Research Brief: Latest ALS GWAS Points to Loci on Chromosomes 9, 19

Paper Citations

  1. . Connecting TDP-43 Pathology with Neuropathy. Trends Neurosci. 2021 Jun;44(6):424-440. Epub 2021 Apr 5 PubMed.
  2. . Loss of Nuclear TDP-43 Is Associated with Decondensation of LINE Retrotransposons. Cell Rep. 2019 Apr 30;27(5):1409-1421.e6. PubMed.
  3. . Truncated stathmin-2 is a marker of TDP-43 pathology in frontotemporal dementia. J Clin Invest. 2020 Nov 2;130(11):6080-6092. PubMed.
  4. . Stathmin-2: adding another piece to the puzzle of TDP-43 proteinopathies and neurodegeneration. J Clin Invest. 2020 Nov 2;130(11):5677-5680. PubMed.

External Citations

  1. New York Genome Center (NYGC) ALS Consortium
  2. Mayo Clinic Brain Bank

Further Reading


  1. . UNC13A polymorphism contributes to frontotemporal disease in sporadic amyotrophic lateral sclerosis. Neurobiol Aging. 2019 Jan;73:190-199. Epub 2018 Sep 27 PubMed.

Primary Papers

  1. . Common ALS/FTD risk variants in UNC13A exacerbate its cryptic splicing 2 and loss upon TDP-43 mislocalization. bioRxiv. April 4, 2021.
  2. . TDP-43 represses cryptic exon inclusion in FTD/ALS gene UNC13A. bioRxiv. April 04, 2021.