Expanded C9ORF72 hexanucleotide repeats cause many cases of amyotrophic lateral sclerosis and frontotemporal dementia, but exactly how these expansions work their mischief remains a mystery. Researchers have proposed several different pathogenic mechanisms, including that aberrant dipeptide repeat (DPR) proteins made from the expansion interfere with mRNA splicing. In the June 13 Cell Reports, researchers led by Robin Reed at Harvard Medical School describe how this could happen. Using in vitro assays, they found that DPRs prevented spliceosome assembly through interactions with one portion of this splicing complex, the U2 small nuclear ribonucleoprotein (snRNP) particle. In cells, C9ORF72 DPRs caused the U2 complex to mislocalize to the cytoplasm, away from splicing sites in the nucleus. Subsequent bioinformatics analyses suggested that disruption of U2 function accounts for about 40 percent of the mis-splicing seen in C9ORF72 patient cells, Reed noted.

The data for the mechanism are compelling, but it remains unclear how much it contributes to neurotoxicity, noted Johnathan Cooper-Knock at the University of Sheffield, England, in an email to Alzforum. He was not involved in the research. “One way to address this might be to examine its correlation with clinical phenotype,” he suggested (see full comment below).

Nuclear Breakout.

In control motor neurons (left), the U2 splicing complex (red) occupies the nucleus, but in C9ORF72 motor neurons (right), much of it lurks in cytoplasm. [Courtesy of Cell Reports, Yin et al.]

Previous research had found that C9ORF72 expansions caused mis-splicing in cultured cells (see Aug 2014 news; Conlon et al., 2016). In one study, the degree of this mis-splicing correlated with faster ALS progression (Cooper-Knock et al., 2015). The papers proposed different explanations for how mis-splicing occurred, however, from dipeptide repeat proteins physically gumming up the splicing machinery in the nucleolus to C9ORF72-expanded RNAs sequestering crucial splicing factors such as HNRNPH.

To try to nail down the mechanism, first author Shanye Yin used a cell-free system developed in the Reed lab that combined nuclear extract from HeLa cells with a DNA template for the fly Ftz gene. In this assay, the Ftz gene is transcribed by RNA polymerase and the resulting RNA is then spliced to yield mature transcript. However, when the authors added a 20-repeat length of the toxic DPRs glycine-arginine (GR) or proline-arginine (PR), splicing ground to a halt. The effectiveness of the block depended on the dose of DPR. Further analysis showed that in the presence of DPRs, the spliceosome failed to assemble properly. C9ORF72 RNA, by contrast, had no effect on splicing in this system.

The authors next searched for proteins that associated with GR and PR using pull-down assays, and identified these interactors by mass spectrometry. The most common were components of the U2 snRNP, although other proteins were present as well. The list of interacting proteins largely matches that seen in previous proteomic screens of DPRs, noted Paul Taylor at St. Jude Children’s Research Hospital in Memphis, Tennessee. “That gives us confidence in the data,” he told Alzforum.

The findings suggested a direct interaction of DPRs with the splicing machinery in vitro. Would the same thing happen in patients? The authors examined motor neurons made from induced pluripotent stem cells of people who carried the C9ORF72 expansion. They were surprised to find U2 snRNP lingering outside the nucleus in about half these cells (see image above). It was unclear if the complex leaked out of the nucleus, or simply never made it inside in the first place, Reed noted. U2 snRNP is assembled in the cytoplasm, and faulty transport across the nuclear membrane has quickly become a hallmark of C9ORF72 disease, with many nuclear proteins, such as TDP43 and FUS, getting stuck en route (Jan 2010 newsAug 2015 news). DPRs are known to build up in the cytoplasm, suggesting they could be sequestering the U2 snRNP there. Overexpressing PR in HeLa cells also caused U2 snRNP to loiter in cytoplasm, supporting a direct causal role for these peptides in U2 mislocalization. This was specific to U2, as other components of the splicing apparatus stayed confined to the nucleus.

The authors next wondered how big a role U2 mis-splicing might play in disease. U2 snRNP normally helps splice about one-quarter of all transcripts (see Kfir et al., 2015). Analyzing published RNA expression data from C9ORF72 cerebella and frontal cortices (Prudencio et al., 2015), the authors found that the mis-spliced mRNA was enriched for U2-dependent transcripts. Altogether, U2 snRNP may account for nearly half of the mis-splicing in these cells, they calculated. They saw a much smaller role for HNRNPH-dependent splicing. Notably, in brain samples from sporadic ALS cases, U2-dependent transcripts were no more likely to be mis-spliced than any others, indicating this mechanism is specific for C9ORF72 expansions.

In theory, mis-splicing could lead to toxicity by knocking out essential proteins. In support of this, the authors found numerous transcripts involved in mitochondrial function and gene expression among the disrupted set. Both of these processes are known to be perturbed in ALS.

However, DPRs sequester other proteins as well, which may play an equally important role in damaging cell function, Taylor noted. Many of the proteins bound by DPRs contain low-complexity domains. Such proteins are now believed to be crucial in assembling many cellular organelles (Oct 2015 webinar; May 2017 conference news).—Madolyn Bowman Rogers

Comments

Make a Comment

Comments on this content

  1. Toxicity of dipeptide repeat proteins (DPRs) has been repeatedly demonstrated in models of disease associated with G4C2-repeat expansion of C9ORF72. The question of how that toxicity is mediated has been harder to elucidate but several lines of evidence have converged on disruption of mRNA splicing. In this paper, the group of Robin Reed goes a step further and attempts to define a specific mechanism. They propose that a key event is disturbed function of the U2 snRNP complex. Several observations support this hypothesis: demonstration of DPR binding to U2 snRNP components in vitro; cytoplasmic mislocalization of U2 snRNP components in C9ORF72 cell models; and enrichment of U2-mediated splicing events in aberrant splicing observed in C9ORF72 models. This is very interesting and provides a number of testable hypotheses.

    However, as the authors acknowledge, there are some limitations to their data, for example, splicing assays shown depend on supra-physiological concentrations of DPRs. Most pointedly, although U2-mediated splicing events are enriched in C9ORF72 models, the majority of aberrant splicing in these models is U2-independent. Similarly, although U2 snRNP complex components are enriched in DPR binding partners, the majority of binding partners (including other splicing proteins) are not components of the U2 snRNP.

    In short, the authors provide evidence for a mechanism but it does not appear to capture all of the interactions they describe, and the relative importance to neurotoxicity is uncertain. Addressing this is important before commencing development of therapeutics. One way to do this might be to examine correlations with clinical phenotype. We have previously shown that splicing error rate correlates with disease severity in C9ORF72-patient derived cells (Cooper-Knock et al., 2015); it would be fascinating to know if this correlation was strengthened when splicing events were limited to U2-mediated events.

    The final observation I would make, which applies to more than just this study, is that currently (given the tools available), abundance of the DPRs studied—PR and GR—does not correlate with neurodegeneration in patient postmortem tissue (Davidson et al., 2016) despite their toxicity in cell and animal models. Moreover, GR and PR are relatively absent from important populations such as postmortem motor neurons in ALS patients. This is an important unsolved puzzle for the field. 

    References:

    . C9ORF72 GGGGCC Expanded Repeats Produce Splicing Dysregulation which Correlates with Disease Severity in Amyotrophic Lateral Sclerosis. PLoS One. 2015;10(5):e0127376. Epub 2015 May 27 PubMed.

    . Neurodegeneration in frontotemporal lobar degeneration and motor neurone disease associated with expansions in C9orf72 is linked to TDP-43 pathology and not associated with aggregated forms of dipeptide repeat proteins. Neuropathol Appl Neurobiol. 2016 Apr;42(3):242-54. Epub 2015 Dec 7 PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. C9ORF72 Killer Dipeptides Clog the Nucleolus
  2. TDP-43: Modified and On the Move
  3. ALS Gene Repeats Obstruct Traffic Between Nucleus and Cytoplasm
  4. Protein Liquid-Liquid Phase Transitions: The Science Is About to Gel

Webinar Citations

  1. Fluid Business: Could “Liquid” Protein Herald Neurodegeneration?

Paper Citations

  1. . The C9ORF72 GGGGCC expansion forms RNA G-quadruplex inclusions and sequesters hnRNP H to disrupt splicing in ALS brains. Elife. 2016 Sep 13;5 PubMed.
  2. . C9ORF72 GGGGCC Expanded Repeats Produce Splicing Dysregulation which Correlates with Disease Severity in Amyotrophic Lateral Sclerosis. PLoS One. 2015;10(5):e0127376. Epub 2015 May 27 PubMed.
  3. . SF3B1 association with chromatin determines splicing outcomes. Cell Rep. 2015 Apr 28;11(4):618-29. Epub 2015 Apr 16 PubMed.
  4. . Distinct brain transcriptome profiles in C9orf72-associated and sporadic ALS. Nat Neurosci. 2015 Aug;18(8):1175-82. Epub 2015 Jul 20 PubMed.

Further Reading

Primary Papers

  1. . Evidence that C9ORF72 Dipeptide Repeat Proteins Associate with U2 snRNP to Cause Mis-splicing in ALS/FTD Patients. Cell Rep. 2017 Jun 13;19(11):2244-2256. PubMed.