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


Please login to recommend the paper.


  1. These papers convincingly argue that toxicity due to C9ORF72 GGGGCC repeat expansions arises from arginine dipeptide repeats (DPRs) interactions with proteins containing low-complexity domain (LCDs) inside ribonucleoprotein granule membrane-less organelles. Intriguingly, many of these LCD proteins themselves form inclusions in, and contain mutations causing, ALS (TDP-43, FUS, hnRNP A2). Using pull-down proteomic approaches, in-cell granule localization, and in vitro reconstitution of liquid droplet and hydrogel models of RNP granules, the authors demonstrate that arginine-rich DPRs interact with low-complexity domains of RNA-binding proteins and other granule components, and even intermediate filament disordered domains. These interactions are shown to alter dynamics of these subcellular membrane-less organelles. Though GGGGCC repeat expansions code for non-arginine peptides as well, Lee et al. demonstrate that arginine is required for granule interactions and toxicity. These papers lead to several critical unanswered mechanistic questions regarding DPRs and ALS.

    1) What is the role of arginine-rich regions and RG/RGG repeats natively present in ALS-associated LCD proteins such as FUS? Our previous work demonstrates that although the low-complexity domain of FUS (residues 1-163), which lacks arginine and other positively charged residues, is alone competent for liquid-liquid phase separation into models of RNP granules, full-length FUS, which includes three RGG repeat regions, phase separates at much lower sub-physiological concentration ranges (Burke et al.,  2015; Oct 2015 news). Additionally, several reports demonstrate that these regions are required for toxicity. Therefore, interactions between RGG regions and low-complexity domains natively present may be outcompeted by arginine-rich polymers and altered by ALS-associated mutations. And yet, the atomic structural details of DPR-LCD interaction are still unknown. It will be critical to conclusively determine where physiological RNP LCDs lie on the continuum between intrinsic disorder and ordered cross-β structure, and how DPRs alter this structure. It will also be a top priority to examine the interaction of DPRs with TDP-43, which can self-assemble via α-helical structure present within its LCD (Conicella et al., 2016Sep 2016 news). 

    2) What is the mechanism of the observed toxicity and the suppressor/enhancer phenotypes observed? Given that several LCD-containing proteins that interact with arginine DPRs are included in the list of modifiers, it will be exciting to determine how altering LCD protein levels modulates toxicity.

    3) Finally, does disruption of LCD proteins in RNPs explain not only C9ORF72 ALS, as suggested by these papers, but rather all ALS, including those with no known genetic component? The answer to this question, which may be found using the approaches demonstrated in these papers, will be important for designing strategies to prevent ALS or halt its progression. 


    . Residue-by-Residue View of In Vitro FUS Granules that Bind the C-Terminal Domain of RNA Polymerase II. Mol Cell. 2015 Oct 15;60(2):231-41. Epub 2015 Oct 8 PubMed.

    . ALS Mutations Disrupt Phase Separation Mediated by α-Helical Structure in the TDP-43 Low-Complexity C-Terminal Domain. Structure. 2016 Sep 6;24(9):1537-49. Epub 2016 Aug 18 PubMed.

    View all comments by Nicolas Fawzi
  2. These two papers present a comprehensive addition to our knowledge of the mechanisms of action of C9ORF72 dipeptide repeats. Studying other protein aggregates, investigators identified a plethora of protein interactions, which lead to the general consensus that the key to preventing toxicity caused by protein aggregates is to prevent their formation or to reduce their abundance. The original set of manuscripts on the C9ORF72 dipeptide repeats suggested that interactions with nucleoli or the nuclear pores were a main target of action. The current manuscripts confirm these interactions, but convey the much more compelling case that the dipeptide repeats interact with many, many proteins within the cell, potentially causing injury through a multiplicity of mechanisms.

    View all comments by Benjamin Wolozin
  3. Since the discovery of C9ORF72 hexanucleotide expansions in patients with ALS and frontotemporal dementia (Dejesus-Hernandez et al., 2011; Renton et al., 2011), controversy has ensued about the cellular mechanisms that drive neurodegeneration in this patient cohort (Gitler and Tsuiji, 2016). While some data suggest that haploinsufficiency could contribute to disease (O'Rourke et al., 2016; Sivadasan et al., 2016), the vast majority of studies instead support a dominant gain of function as the major causal player in disease pathogenesis. Cell biologists have been roughly divided into two camps: one that suggests that transcribed C9ORF72 RNA begets dysfunction (Zhang et al., 2015), whereas the other argues that unconventionally translated dipeptides give rise to disease (Zu et al., 2011). Initial mammalian cell culture studies from Steven McKnight’s lab convincingly demonstrated that exogenously applied arginine-rich dipeptides are sufficient to kill cells, likely via dysfunction of the nucleolus (Kwon et al., 2014). Screens from Aaron Gitler and Paul Taylor’s labs further suggested that nuclear import/export could be impaired by arginine-rich dipeptides, but not by other dipeptides produced from the C9ORF72 locus (Freibaum et al., 2015; Jovičić et al., 2015), although the Taylor study did not rule out a role for expanded RNA transcripts in driving toxicity.

    This new Cell paper from Paul Taylor and colleagues is an elegant piece of work that combines proteomics, Drosophila screening, in vitro biochemistry, and live cell dynamics to show that arginine-rich dipeptides “poison” a vast array of membraneless organelles that arise by liquid-liquid phase separation from the nucleoplasm and cytoplasm (Brangwynne et al., 2009; Weber and Brangwynne, 2012). This seminal paper can be appreciated from two different angles:

    1. From a therapeutic perspective, because it suggests that drug development should aim to reduce arginine-rich dipeptide production in patients with C9ORF72 hexanucleotide repeat expansions.
    2. From a basic cell biology standpoint, because it gives further credence to the hypothesis that electrostatics and multivalent blocked-charge motifs are of critical importance to the assembly and regulation of diverse membraneless organelles, e.g., nucleoli, splicing speckles, Cajal bodies, stress granules, paraspeckles (Brangwynne et al., 2015; Nott et al., 2015; Pak et al., 2016). 

    While I applaud the importance of the findings from a cell biology perspective (I know the term “tour de force” is used too often by scientific commentators, but nevertheless, I think the phrase is applicable in this case), I wish to add some caveats regarding the relevance to disease pathogenesis, which I believe are lacking in the Alzforum perspective. First, I must emphasize that the authors do not argue that arginine-rich dipeptides are the only contributing factor to C9ORF-ALS/FTD pathogenesis.

    It is now quite clear from numerous studies that large amounts of positively charged dipeptides are unequivocally “bad” for cells. This is likely due to the fact that arginine-rich dipeptides compete with endogenous RGG motif-containing proteins for electrostatic interactions with negatively charged and phosphorylated proteins enriched in membraneless organelles. However, it is not clear to me how accurately current cell and organismal models recapitulate critical features of C9ORF-ALS/FTD. Exhaustive immunohistochemical studies have shown neuronal loss correlates with TDP-43 pathology but not dipeptide pathology (Mackenzie et al., 2015). Moreover, unlike all dipeptide models, most human dipeptide pathology is observed in the cytoplasm and not the nucleus, (Mackenzie et al., 2015). Further, poly-GA and poly-GP inclusions are far more abundant than poly-PR or poly-GR accumulations (Mackenzie et al., 2015). Finally, when dipeptides are observed in the nucleus, they appear to localize to a perinucleolar compartment and not the granular component of the nucleolus (Schludi et al., 2015; Vatsavayai et al., 2016). With these pathological observations in mind, I think it is important that we continue to consider all possible mechanisms (haploinsufficiency, RNA toxicity, arginine dipeptide toxicity, GA dipeptide toxicity) in the pathogenesis of C9ORF-ALS/FTD. Going forward, some important topics/questions come to mind that will be essential to explore for this rapidly evolving research field (i.e., the relationship between physiological phase separation and pathological protein aggregation and neurodegeneration):

    1. To what degree does each proposed pathological mechanism give rise to TDP-43 pathology? Do dipeptides, C9ORF72 haploinsufficiency, or RNA toxicity directly drive the accumulation of TDP-43 in the cytoplasm of neurons? Are these upstream factors the “trigger” and TDP-43 the “bullet” (i.e., TDP-43 is initially induced by C9ORF72-related pathology to form self-propagating prions that then spread throughout the brain to drive progressive neurodegeneration (Sanders et al., 2016)), or is TDP-43 accumulation merely a correlate irrelevant to disease pathogenesis? It is interesting that a recent paper from Bill Seeley’s lab shows that neurodegeneration can occur in the absence of overt TDP-43 pathology, at least in certain patients (Vatsavayai et al., 2016). 
    2. Why is TDP-43 so prone to forming amyloid-like cytoplasmic aggregates, relative to other RNA-binding proteins with prion-like and low-complexity domains? For example, G3BP1 is thought to be a “condensate” that allows mRNPs to self-associate into membraneless organelles called stress granules (Kedersha et al., 2013; 2016), but does not accumulate in large fibrils in ALS/FTD.
    3. If arginine-rich dipeptides are killing neurons, how do they do so in patients? Pathology studies do not show massive accumulations of these dipeptides in nucleoli and other nuclear bodies. Is it possible that this is simply a detection problem (i.e., antibodies are not sensitive enough or epitopes are buried)? Perhaps prolonged, low-level expression of dipeptides slowly alters membraneless organelle dynamics until “catastrophes” occur, whereby proteins like TDP-43 undergo a liquid-to-solid phase transition and become a self-propagating amyloid (or prion) (Sanders et al., 2016; Weber and Brangwynne, 2012). 
    4. If proteins in membraneless organelles are, in fact, in a precarious biophysical state and are prone to undergoing liquid-to-solid phase transitions (Weber and Brangwynne, 2012), why do we so rarely observe nuclear accumulations of solid fibril-like structures in patient brains relative to cytoplasmic aggregates? What is allowing the fluidity of these structures to be maintained over years and decades?
    5. With respect to the current paper from Taylor’s group, why does knockdown of ostensibly similar proteins lead to divergent outcomes (enhancers versus suppressors)? On cursory examination, there does not appear to be any logic to this at the level of amino acid composition (e.g., isoelectric point). This was the most fascinating aspect of the paper to me, as it suggests that there is still a wealth of information to be discovered from the study of proteins that form these organelles. It is not as simple as “electrostatics.” Moreover, I wish to emphasize the incredible percentage of modifiers from the biased screen (85 percent!). Nevertheless, it would have been nice to see a larger control panel of genes (e.g., generic proteins or even better yet, RNA-binding proteins without LCS domains) to get a good sense of the specificity of enhancer/suppressor phenotypes to the proteomic hits. Two controls were used, V60100 and w118, which showed ~25 percent and ~48 percent viability, respectively. Thus, it appears that the genetic system used in the paper may be especially prone to perturbation. A weakness of the study was that the cutoff to define an enhancer or suppressor was somewhat arbitrary based on the number of controls. 
    6. What accounts for the differences between cell culture/organismal models and patients with regard to sub-nuclear accumulations of arginine-rich dipeptides? Is it length or concentration dependent? This is a fascinating and vexing problem!
    7. If it is true that C9ORF72 dipeptides disrupt nuclear bodies and lead to neuronal dysfunction, why is it that pathology spreads in a hierarchical and stereotyped manner in C9ORF-ALS/FTD patients? The most parsimonious explanation for this observation is the release and uptake of a toxic factor (e.g., a prion) in a neuronal network (Sanders et al., 2016). Do dipeptides act like prions? The vast majority of studies do not as yet support this hypothesis, although one recent study demonstrated transcellular spread of dipeptides, but not “seeding” in the recipient cells (Westergard et al., 2016). 
    8. Finally, I think many questions are raised about the dynamic interplay between these numerous RNP bodies in the cells. Future work must determine what regulates shuttling of individual proteins from one compartment to the next, and how individual proteins are targeted to specific organelles despite similar amino acid compositions (Nott et al., 2015). 

    Like any great paper, these two lead to more questions than they answer. Needless to say, it is an exciting time for the field fof phase separation and neurodegeneration!


    . Germline P granules are liquid droplets that localize by controlled dissolution/condensation. Science. 2009 Jun 26;324(5935):1729-32. Epub 2009 May 21 PubMed.

    . Polymer physics of intracellular phase transitions. Nature Physics 11, 899–904. (2015)

    . Expanded GGGGCC hexanucleotide repeat in noncoding region of C9ORF72 causes chromosome 9p-linked FTD and ALS. Neuron. 2011 Oct 20;72(2):245-56. Epub 2011 Sep 21 PubMed.

    . GGGGCC repeat expansion in C9orf72 compromises nucleocytoplasmic transport. Nature. 2015 Sep 3;525(7567):129-33. Epub 2015 Aug 26 PubMed.

    . There has been an awakening: Emerging mechanisms of C9orf72 mutations in FTD/ALS. Brain Res. 2016 Sep 15;1647:19-29. Epub 2016 Apr 6 PubMed.

    . Modifiers of C9orf72 dipeptide repeat toxicity connect nucleocytoplasmic transport defects to FTD/ALS. Nat Neurosci. 2015 Sep;18(9):1226-9. PubMed.

    . Stress granules and cell signaling: more than just a passing phase?. Trends Biochem Sci. 2013 Oct;38(10):494-506. Epub 2013 Sep 10 PubMed.

    . G3BP-Caprin1-USP10 complexes mediate stress granule condensation and associate with 40S subunits. J Cell Biol. 2016 Mar 28;212(7):845-60. PubMed.

    . 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.

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

    . Quantitative analysis and clinico-pathological correlations of different dipeptide repeat protein pathologies in C9ORF72 mutation carriers. Acta Neuropathol. 2015 Dec;130(6):845-61. Epub 2015 Sep 15 PubMed.

    . Phase transition of a disordered nuage protein generates environmentally responsive membraneless organelles. Mol Cell. 2015 Mar 5;57(5):936-47. PubMed.

    . C9orf72 is required for proper macrophage and microglial function in mice. Science. 2016 Mar 18;351(6279):1324-9. PubMed.

    . Sequence Determinants of Intracellular Phase Separation by Complex Coacervation of a Disordered Protein. Mol Cell. 2016 Jul 7;63(1):72-85. PubMed.

    . A hexanucleotide repeat expansion in C9ORF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron. 2011 Oct 20;72(2):257-68. Epub 2011 Sep 21 PubMed.

    . Prions and Protein Assemblies that Convey Biological Information in Health and Disease. Neuron. 2016 Feb 3;89(3):433-48. PubMed.

    . Distribution of dipeptide repeat proteins in cellular models and C9orf72 mutation cases suggests link to transcriptional silencing. Acta Neuropathol. 2015 Oct;130(4):537-55. Epub 2015 Jun 18 PubMed.

    . C9ORF72 interaction with cofilin modulates actin dynamics in motor neurons. Nat Neurosci. 2016 Dec;19(12):1610-1618. Epub 2016 Oct 10 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.

    . Getting RNA and protein in phase. Cell. 2012 Jun 8;149(6):1188-91. PubMed.

    . Cell-to-Cell Transmission of Dipeptide Repeat Proteins Linked to C9orf72-ALS/FTD. Cell Rep. 2016 Oct 11;17(3):645-652. PubMed.

    . The C9orf72 repeat expansion disrupts nucleocytoplasmic transport. Nature. 2015 Sep 3;525(7567):56-61. Epub 2015 Aug 26 PubMed.

    . Non-ATG-initiated translation directed by microsatellite expansions. Proc Natl Acad Sci U S A. 2011 Jan 4;108(1):260-5. Epub 2010 Dec 20 PubMed.

    View all comments by David Sanders

Make a Comment

To make a comment you must login or register.