As the most common genetic cause of amyotrophic lateral sclerosis and frontotemporal dementia (ALS/FTD), repeat expansions in the C9ORF72 gene spew a poisonous brew of abnormal RNAs and small dipeptide repeat (DPR) proteins. In a new mouse model, researchers led by Leonard Petrucelli at the Mayo Clinic in Jacksonville, Florida, reveal that one of those proteins, the poly glycine-arginine (GR) dipeptide, kills neurons by targeting ribosomes and quenching new protein synthesis. The investigators also spotted poly(GR) and ribosomal proteins together in brain tissue from people with C9-ALS/FTD, suggesting that finding ways to break up the toxic union could offer potential therapies for the disease. In cells and in the mice, though not in human samples, poly(GR) also localized to stress granules. The study appeared June 25 in Nature Medicine.

  • GR dipeptides from C9ORF72 ALS/FTD cause neurodegeneration in mice.
  • Poly(GR) inhibits protein translation and keeps stress granules from dissolving.
  • Patient tissue also shows poly(GR) binding ribosomes.

“This new paper reports the first mouse model expressing specifically poly(GR), which is an exciting step forward for the field,” said Adrian Isaacs, University College, London (full comment below). Paul Taylor of St. Jude Children’s Research Hospital in Memphis, Tennessee, agreed. “Arginine-containing poly-dipeptides (PR and GR) produced in C9-ALS/FTD are exquisitely toxic in a wide variety of model systems. … This paper adds important new insight to this emerging story by showing that, in transgenic mice, expressing poly-GR at levels comparable to, or lower than, levels observed in human C9-ALS/FTD is sufficient to drive neurodegeneration,” Taylor wrote to Alzforum (see full comment below).

The study extends recent findings from Dieter Edbauer at Ludwig-Maximilians University, Munich (Jun 2018 news). Edbauer’s group discovered that poly(GR) mainly targets ribosomes in primary neurons and in brain samples from C9-FTD patients. “It will be interesting to determine whether boosting translation would prevent neuronal death, or even worsen toxicity through enhanced DPR production,” Edbauer wrote in an email to Alzforum (see full comment below).

Expansion of the six-nucleotide repeat GGGGCC in C9ORF72 gives rise to two abnormal RNAs, which encode five DPR proteins: GR, proline-arginine (PR), glycine-alanine (GA), proline-alanine, and glycine-proline. All accumulate in ALS/FTD. Previously, Petrucelli’s group had generated a mouse model expressing 66 of the GGGGCC repeats from an adenovirus vector they injected into the newborn mouse brain (May 2015 news). Those mice recapitulated key features of ALS pathology, including DPR-containing aggregates, cytoplasmic inclusions harboring phosphorylated TDP-43, neurodegeneration, and behavioral changes.

In the new study, first author Yong-Jie Zhang used a similar approach to express just the poly(GR) DPR. Based on studies in yeast, flies, and cells, arginine-containing peptides are the most toxic in the C9ORF72 mix, and that held true in the mice. Injection of a virus encoding green fluorescent protein tagged with 100 GR repeats resulted in diffuse cytoplasmic poly(GR) expression. The mice expressed levels of poly(GR) equivalent to some patient samples and the GGGGCC mice, but had almost none of the compact cytoplasmic inclusions seen in those instances. This is consistent with previous work indicating that poly(GR) needs other DPR proteins to promote its own aggregation (Yang et al., 2015). 

But even without aggregating, poly(GR) produced severe neurodegeneration. Already at six weeks of age, neuroinflammation markers in the cortices and hippocampi of GFP-(GR)100 mice were elevated. By six months of age, the brains of GR mice weighed just over half those of GFP-expressing controls. The GR mice had but one-third as many cortical neurons, and one-quarter as many CA1-CA3 hippocampal neurons as did controls. They developed significant motor and cognitive deficits: They travelled less in an open field test, were clumsier when walking on a rod or hanging from a wire, and were less able to associate a noise cue with imminent foot shock in a fear conditioning test.

Consistent with the prevailing view that GR repeat peptides are more toxic than GA, the new GR mice had earlier onset and more severe neurodegeneration than mice expressing poly(GA) (Mar 2016 news). 

But how does it happen? In cells, poly(GR) DPRs bind ribosome proteins and disrupt translation (Kanekura et al., 2016; Oct 2016 news; Lopez-Gonzalez et al., 2016). The researchers observed the same in mice. Using immunofluorescence, they saw the distribution of poly(GR) overlap with S6, L21, and S25 ribosomal proteins, as well as the translation initiation factor eIF3η. The effect was selective for GR, as the same pairing was absent in mice expressing poly(GA). Importantly, both diffuse and aggregated poly(GR) co-localized with S6, L21, and eIF3η in postmortem brain tissue from C9-FTD/ALS patients, as well as in mice expressing the GGGGCC repeat.

GRinding to a Halt. Mouse cortical cells expressing poly(GR) (green, left panel) are deficient in protein synthesis, as revealed by lower incorporation of the protein translation tag puromycin (red cells with arrows, center panel). [Courtesy of Zhang et al., 2018, Nature Medicine.]

A transcriptome analysis of brain tissue from the mice revealed expression changes in two main gene clusters: One involved the immune response, and the other spanned the activity and constituents of ribosomes and protein translation. Expression of ribosomal proteins was up in GR mice, presumably to compensate for low ribosomal activity. Indeed, GR dipeptide repeat protein inhibited new protein synthesis in HEK293T cells, as measured by incorporation of puromycin. The investigators also detected this in the mice, where cells expressing GR incorporated less puromycin than cells without (see image). Poly(GR) even shut down the special translational pathway cells use to produce DPRs from the C9ORF72 repeat RNA (Jan 2018 news), suggesting that GR might negatively regulate its own production.

Another likely poly(GR) target is the stress granule, a dynamic collection of RNA and protein that forms in cells in distress. In cells, poly(GR) overexpression promotes stress granule formation and impedes their dissolution (Oct 2016 news; Tao et al., 2015). In mice, diffuse poly(GR) did not increase formation of stress granules like the GR-containing aggregates could, but when stress granules were induced by way of heat shock, diffuse poly(GR) was recruited to them and impaired their disassembly. Besides inhibiting protein synthesis, promoting the persistence of SGs could be another way in which poly(GR) induces a state of chronic stress in cells, the authors speculated. They did not find evidence of poly(GR) in stress granules in human ALS/FTD brain. The reasons are unclear, Petrucelli told Alzforum.

The mice are notable for the pathology they do not show, Petrucelli said. For example, neurons degenerated even when there was no TDP-43 accumulation. It’s unclear how C9ORF72 expansions are linked to cytosolic accumulation of phosphorylated TDP-43, and Petrucelli said his group intends to use mouse models to systematically query each of the DPRs to see which elicit neurotoxicity, and how. The story will likely be complicated, as RNA appears to foment TDP-43 aggregation and TDP-43 itself was recently linked to disrupted protein synthesis (May 2015 news; Feb 2017 news). Petrucelli wants to explore drivers of poly(GR) aggregation in vivo, and whether poly(GA) or other DPRs cooperate in that process.

The investigators also found no evidence for nuclear dysfunction, which surprised Ben Wolozin of Boston University. “Since poly(GGGGCC) pathology is associated with dysfunction of nuclear and nuclear pore biology in other animal models, the absence of such dysfunction in the poly(GR) models once again points to the mechanistic specificity by which poly(GR) causes damage, and suggests that nuclear pore pathology does not inevitably occur during neurodegeneration,” he wrote to Alzforum (full comment below).—Pat McCaffrey

Comments

  1. This new paper reports the first mouse model expressing specifically poly(GR), which is an exciting step forward for the field. This was achieved by expressing 100 GFP-tagged poly(GR) repeats in the mouse brain using AAV. The mice show diffuse cytoplasmic GR staining and profound neurodegeneration accompanied by behavioral and motor deficits.

    A really nice aspect of the paper is the ability to compare to both GGGGCC-repeat mice and polyGA mice. For instance, poly(GR) alone was not able to induce the TDP-43 inclusions observed in the GGGGCC-repeat mice (and neither was poly(GA)). Therefore it seems that TDP-43 pathology in the GGGGCC-AAV mice is induced by either:

    1.  poly(GR) inclusions rather than the diffuse poly(GR) observed in the model,
    2. a distinct DPR (poly(GP), poly(PR), and poly(AP) are left to examine in this AAV model),
    3. a synergistic effect of two or more different DPRs,
    4. a synergistic effect requiring both DPRs and repeat RNA,
    5. repeat RNA alone, or
    6. lower DPR expression is required to tease out TDP-43 pathology, whereas higher expression may lead to direct toxicity without time to invoke TDP-43 changes.

    These are important questions to now figure out.

    A central message of the paper is that poly(GR) but not poly(GA) associated with ribosomal subunits and inhibited translation, which is consistent with previous data. Whether translation inhibition is due to direct ribosome binding, the altered ribosomal gene expression also reported, or previously suggested mechanisms such as RNA binding, remains an open question.

    Either way, these data now strongly implicate altered translation as an important factor in C9ORF72 pathogenesis. A critical next step will be to determine whether it is possible to reduce this impaired translation and if so whether such an intervention can rescue poly(GR)-induced neurodegeneration.

  2. This paper places another important puzzle piece. Arginine-containing poly-dipeptides (PR and GR) produced in C9-ALS/FTD are exquisitely toxic in a wide variety of model systems, including HeLa cells (Kwon et al., 2014), cultured neurons (Wen et al., 2014), and Drosophila (Mizielinska et al., 2014), whereas the other polypeptides produced in C9-ALS/FTD are either innocuous (PA and PG) or only mildly toxic at high levels (GA).

    One important basis for PR and GR toxicity in these systems is infiltration and disturbance of structures that arise by phase transitions. This disturbance includes stress granules and impairs translation, but is not limited to these organelles (Lee et al., 2016). However, it has been unclear how the quantities of poly-dipeptides expressed in these model systems relate to the quantities found in human C9-ALS/FTD. An important recent human study found that regional levels of GR in brains of C9-ALS-FTD patients correlates with neurodegeneration and burden of TDP-43 pathology (Saberi et al., 2018); this further implicates GR in disease but causation is not easily inferred from correlative postmortem studies.

    This paper adds important new insight to this emerging story by showing that, in transgenic mice, expressing poly(GR) at levels comparable to, or lower than, levels observed in human C9-ALS/FTD is sufficient to drive neurodegeneration.

    It is important to note, however, that these animals did not develop significant TDP-43 pathology, a hallmark of C9-ALS/FTD and most other sporadic and familial forms of ALS-FTD. Perhaps a second factor is needed to initiate this aspect of the disease phenotype, or a longer disease duration is required for significant accrual of TDP-43 pathology.

    References:

    . Poly-dipeptides encoded by the C9orf72 repeats bind nucleoli, impede RNA biogenesis, and kill cells. Science. 2014 Sep 5;345(6201):1139-45. Epub 2014 Jul 31 PubMed.

    . Antisense proline-arginine RAN dipeptides linked to C9ORF72-ALS/FTD form toxic nuclear aggregates that initiate in vitro and in vivo neuronal death. Neuron. 2014 Dec 17;84(6):1213-25. PubMed.

    . C9orf72 repeat expansions cause neurodegeneration in Drosophila through arginine-rich proteins. Science. 2014 Sep 5;345(6201):1192-4. Epub 2014 Aug 7 PubMed.

    . C9orf72 Dipeptide Repeats Impair the Assembly, Dynamics, and Function of Membrane-Less Organelles. Cell. 2016 Oct 20;167(3):774-788.e17. PubMed.

    . Sense-encoded poly-GR dipeptide repeat proteins correlate to neurodegeneration and uniquely co-localize with TDP-43 in dendrites of repeat-expanded C9orf72 amyotrophic lateral sclerosis. Acta Neuropathol. 2018 Mar;135(3):459-474. Epub 2017 Dec 1 PubMed.

  3. Zhang et al. report neurodegeneration and behavioral changes in the first poly(GR) mouse model using their AAV-based expression system. Strikingly, these mice show no TDP-43 pathology, suggesting that DPRs also trigger TDP-43-independent pathways of neurotoxicity.

    Unfortunately, this leaves the mechanistic link between the C9ORF72 repeat expansion and TDP-43 pathology found in patients still unresolved. Consistent with our recent report in primary neurons (Hartmann et al., 2018), Zhang et al. conclude that poly(GR) mainly affect ribosomal function in mice and find no evidence for impaired nucleocytoplasmic transport in vivo. Unfortunately, they did not analyze splicing changes in their RNAseq data (compare (Kwon et al., 2014). 

    They extend previous observations that cytoplasmic poly(GR) aggregates co-localize with stress granule markers in mice, but do not provide data for C9ORF72 patient tissue. In our attempt to validate neuronal poly-GR/PR interactors in C9ORF72 brains, we could not detect classical stress granule markers such as TIAR and G3BP2 in patients, and apart from ribosomes only detected co-localization with STAU2, which has also been linked to stress granules.

    Taken together, chronic impairment of translation by poly(GR) may contribute to neurodegeneration in C9ORF72 patients. It will be interesting to determine whether boosting translation would prevent neuronal death or even worsen toxicity through enhanced DPR production.

    References:

    . Proteomics and C9orf72 neuropathology identify ribosomes as poly-GR/PR interactors driving toxicity. Life Science Alliance. 16 May 2018. DOI: 10.26508/lsa.201800070

    . Poly-dipeptides encoded by the C9orf72 repeats bind nucleoli, impede RNA biogenesis, and kill cells. Science. 2014 Sep 5;345(6201):1139-45. Epub 2014 Jul 31 PubMed.

  4. The manuscript by Zhang et al. provides a new model for the study of C9ORF72, and in the process provides useful insight into common pathways that appear to mediate neurodegeneration in ALS and related disorders. Previously, Petrucelli’s group showed that strong expression of (G4C2)66 in the mouse brain produces a model with abundant dipeptide repeats, phospho-TDP-43 inclusions, and neurodegeneration. DPRs, though, come in six flavors, with poly(GR) being best correlated with neurodegeneration. In the current manuscript, Petrucelli’s group has specifically created mice expressing GFP-poly(GR)100 delivered by AAV1, which preferentially infects neurons. They observe that the mouse brains exhibit abundant cytoplasmic poly(GR) as well as exhibiting neurodegeneration and inflammation.

    The nature of the pathology is highly informative, and provides key insights into pathways that contribute to the neurodegeneration. The fact that poly(GR) ​correlates with neurodegeneration is important, since poly(GA) doesn’t correlate with neurodegeneration, and supports accumulating data that dipeptides with arginines are particularly toxic. Arginines are notable for being one of the classic amino acids that are abundant in low-complexity domains of proteins (which promote liquid-liquid phase separation). However, poly(GA) does appear to play a part in the aggregation of poly(GR), which the group shows is recruited to poly(GA) inclusions and co-localizes with poly(GA) inclusions in mice expressing poly(G4C2)100. These data indicate that the type of DPR is critical for toxicity, and that toxicity is quite distinct from aggregation. Another moderate surprise is that toxicity is also separate from aggregation of TDP-43. A fraction of DPR inclusions in poly(G4C2)100 mice co-localize with aggregated phospho-TDP-43; the presence of aggregated phospho-TDP-43 raises the possibility that it plays some role in the ensuing neurodegeneration. However, the absence of TDP-43 inclusions in the poly(GR)100 mice shows that poly(GR)100 is sufficient to cause neurodegeneration without the involvement of TDP-43.

    One of the other pathologies that is surprisingly missing is dysfunction of the nuclear pore and nucleus generally. Since poly(G4C2) pathology is associated with dysfunction of nuclear and nuclear pore biology in other animal models, the absence of such dysfunction in the poly(GR) models once again points to the mechanistic specificity by which poly(GR) causes damage, and suggests that nuclear pore pathology does not inevitably occur during neurodegeneration.

    The next set of insights comes when the group considers which types of proteins co-localize with the poly(GR) inclusions (in the poly(GR)100 mice). They find that the poly(GR) inclusions co-localize with classic markers of stress granules, including TIA1 and EIF3η. This provides support for the hypothesis that the pathophysiology of the DPR inclusions involves stress granules or a related RNA granule. Cell-culture studies show that the poly(GR)100 exerts a particularly strong effect on delaying dispersal of stress granules, which is consistent with a growing body of evidence in the field suggesting that the main effect of mutations in RNA-binding proteins is to slow dispersal/removal of stress granules. 

    Stress granules are thought to be part of the translational stress response, which adapts RNA translation to cope with stress. Co-localization experiments combined with RNAseq experiments show that poly(GR) inclusions co-localize with many ribosomal proteins, including RPS25, which is involved in non-canonical RNA translation. It is important to note that the group finds co-localization with ribosomal proteins from both the large 60S subunit (RPLs) as well as the smaller 40S subunit (RPSs). The presence of both RPL and RPS proteins is consistent with what we have seen in inclusions in brains of models of tauopathies, but differs from the classic view of stress granules, which are thought to contain only RPS subunits. This difference could reflect differences in the regulatory pathways and/or differences between the biology of non-dividing neurons vs. transformed peripheral cell lines.

    Taken together, this work highlights the specific pathophysiological mechanisms of poly-(GR)100, and demonstrates that the pathophysiology of DPRs differ by DPR type. The strong connection between poly(GR) and stress granule (or related) biology also highlights mechanisms shared in common with other neurodegenerative diseases.

  5. In this paper the Petrucelli lab extend their previous work by generating new viral-mediated models of poly-glycine-arginine (GR) expression in mice.

    The group has previously demonstrated that overexpression of sense-direction 66 repeat RNA via intracerebroventricular injection of adeno-associated virus (AAV) leads to a strongly toxic phenotype in mice (Chew et al., 2015). These mice demonstrate sense RNA foci, dipeptide repeat proteins, and nuclear TDP-43 pathology.

    Here, the group overexpressed GFP-tagged, poly(GR)100 repeats alone and observed a strong neuronal toxicity. Interestingly, in contrast to the repeat RNA expressing model, the GR mice do not develop TDP-43 pathology, or poly(GR) aggregates, suggesting that other dipeptide proteins or repeat RNA are required to mediate this aggregative process.

    Based on previous studies that have identified ribosomal proteins as key poly(GR) interacting proteins (Lee et al., 2016; Lin et al., 2016; Boeynaems et al., 2017), the group demonstrated that poly(GR) was able to induce translational repression, and that poly(GR) co-localized with ribosomal proteins in mouse and patient brain. These findings fit well with previous reports that arginine-rich dipeptides like poly(GR) can lead to translational repression (Kanekura et al., 2016; Lee et al., 2016; Hartmann et al., 2018). However the mechanism(s) that cause this translational repression are not fully resolved and may include: binding of ribosomes by dipeptide proteins, induction of stress granules, downregulation of ribosomal proteins, or binding of RNA by arginine containing dipeptides.

    Further work will be required to distinguish these potential mechanisms, although these results strongly suggest that targeting translational repression may be a therapeutic avenue in C9ORF72 ALS/FTD.

    The authors demonstrate that, besides suppressing canonical translation, poly(GR) is capable of also suppressing repeat-associated non-AUG initiated translation (RANT). This could provide a compelling mechanism as to how a seemingly low-abundance protein can induce pathology, as poly(GR) may simultaneously induce a toxic inhibition of translation whilst also preventing its own expression.

    References:

    . Neurodegeneration. C9ORF72 repeat expansions in mice cause TDP-43 pathology, neuronal loss, and behavioral deficits. Science. 2015 Jun 5;348(6239):1151-4. Epub 2015 May 14 PubMed.

    . C9orf72 Dipeptide Repeats Impair the Assembly, Dynamics, and Function of Membrane-Less Organelles. Cell. 2016 Oct 20;167(3):774-788.e17. PubMed.

    . Toxic PR Poly-Dipeptides Encoded by the C9orf72 Repeat Expansion Target LC Domain Polymers. Cell. 2016 Oct 20;167(3):789-802.e12. PubMed.

    . Phase Separation of C9orf72 Dipeptide Repeats Perturbs Stress Granule Dynamics. Mol Cell. 2017 Mar 16;65(6):1044-1055.e5. PubMed.

    . Poly-dipeptides encoded by the C9ORF72 repeats block global protein translation. Hum Mol Genet. 2016 May 1;25(9):1803-13. Epub 2016 Feb 29 PubMed.

    . Proteomics and C9orf72 neuropathology identify ribosomes as poly-GR/PR interactors driving toxicity. Life Science Alliance. 16 May 2018. DOI: 10.26508/lsa.201800070

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Dipeptide Repeats May Hobble Ribosomes in C9ORF72-FTD Patient Brain
  2. First C9ORF72 Mice Mimic Key Pathology, Behavior
  3. C9ORF72 Dipeptide Repeat Not Enough for Motor Neuron Disease
  4. ALS Research ‘Gels’ as Studies Tie Disparate Genetic Factors Together
  5. Stressed-Out Cells Translate C9ORF72 Repeats, Unleash Toxic Peptides
  6. Antisense RNA from C9ORF72 Repeats Is Likely Culprit in Patient Neurons
  7. In New Role for TDP-43, Scientists Say it Controls Protein Synthesis

Research Models Citations

  1. C9ORF72(AAV)(G4C2)66
  2. AAV-GFP–(GR)100

Paper Citations

  1. . FTD/ALS-associated poly(GR) protein impairs the Notch pathway and is recruited by poly(GA) into cytoplasmic inclusions. Acta Neuropathol. 2015 Oct;130(4):525-35. Epub 2015 Jun 2 PubMed.
  2. . Poly-dipeptides encoded by the C9ORF72 repeats block global protein translation. Hum Mol Genet. 2016 May 1;25(9):1803-13. Epub 2016 Feb 29 PubMed.
  3. . Poly(GR) in C9ORF72-Related ALS/FTD Compromises Mitochondrial Function and Increases Oxidative Stress and DNA Damage in iPSC-Derived Motor Neurons. Neuron. 2016 Oct 19;92(2):383-391. Epub 2016 Oct 6 PubMed.
  4. . Nucleolar stress and impaired stress granule formation contribute to C9orf72 RAN translation-induced cytotoxicity. Hum Mol Genet. 2015 May 1;24(9):2426-41. Epub 2015 Jan 9 PubMed.

Further Reading

Primary Papers

  1. . Poly(GR) impairs protein translation and stress granule dynamics in C9orf72-associated frontotemporal dementia and amyotrophic lateral sclerosis. Nat Med. 2018 Aug;24(8):1136-1142. Epub 2018 Jun 25 PubMed.