Much like other RNA-binding proteins, TDP-43 sports a disordered domain at its C-terminal end. Researchers believe these floppy tails enable the proteins to move effortlessly into and out of RNA granules—membraneless organelles that package and process RNA. A paper in the August 18 Structure suggests that for TDP-43, α-helices in the otherwise disordered domain form an intricate lattice that allows RNA granules to separate from the surrounding cytoplasm in a process known as liquid-liquid phase separation. Scientists led by Nicolas Fawzi, Brown University, Providence, Rhode Island, report that mutations associated with amyotrophic lateral sclerosis disrupt this phase transition, making the protein more likely to aggregate instead.

“From people to fish, we have the same exact sequence in this helical region,” Fawzi told Alzforum. “Now we are able to provide some new insight about what that region does.”

Other scientists agreed that the findings represent a significant advance. “Since we know very little about the protein-protein interactions that underlie phase separation, characterizing the contribution of the helix is new and important,” wrote Paul Taylor, St. Jude Children's Research Hospital, Memphis, to Alzforum. Taylor was not involved in the research.

Protein Phase.

TDP-43’s C-terminal helices (pink) form an orderly assembly (top) that allows liquid-liquid phase separation. Mutations (black stars) disrupt this process, promoting protein aggregation instead (bottom). [Image courtesy of Structure, Conicella et al.]

Taylor and others previously had proposed that these disordered regions—also known as low-complexity domains—at the tail end of TDP-43 and other RNA-binding proteins such as FUS and hnRNPA1 facilitate the formation of RNA-protein assemblies such as stress granules (Molliex et al., 2015). Following on that idea, Fawzi’s group proposed that these assemblies formed via transient contacts among the proteins’ C-terminal domains, which remained disordered (Oct 2015 news on Burke et al., 2015). But researchers led by Jianxing Song, National University of Singapore, reported that the C-terminal domain of TDP-43 likely contains an evolutionarily conserved α-helix that is important for function, possibly allowing the protein to embed itself in the cell membrane (Lim et al., 2016). Could that ordered structure be involved in the protein-protein interactions of this otherwise disordered domain? 

To test this, first author Alexander Conicella and colleagues first used NMR spectroscopy and other biophysical techniques to determine that amino acids 321 to 330 of the TDP-43 protein do form an α-helix. Interestingly, upon interacting with a partner helix, the structure became even more structured, with the helix extending to include amino acids 331 to 340. They then demonstrated that in a test tube the TDP-43 C-terminal domain can form liquid droplets reminiscent of RNA-protein assemblies.

Next, Conicella and colleagues found that the α-helix was crucial for this liquid phase separation. If the authors deleted the helical domain or introduced mutations that left it intact but interrupted the helix-helix interactions, no droplets materialized. Likewise, ALS mutations A321G and A321V, which slightly disrupt the helix, limited droplet formation—either they failed to form or they formed less efficiently. In a matter of hours though, these mutated TDP-43 proteins started to clump together into aggregates, which could be seen by differential interference contrast microscopy.

Taken together, the data suggest that in the normal TDP 43, α-helices align precisely so that the proteins interact in an orderly way. However, if a minor change disrupts that interaction, the protein clumps up and forms aggregates.

The work supports the hypothesis that ALS mutations lead to a loss of normal protein function, said Fawzi, though he did not rule out an additional toxic gain of function for the aggregates, as has been proposed previously (see Oct 2015 news). Fawzi noted that this helical interaction might be a good therapeutic target for stabilization by a small molecule. Researchers are pursuing a similar strategy for other protein assemblies such as the tetramers normally formed by the proteins transthyretin or α-synuclein (Dec 2013 news; Apr 2015 conference news). 

Fawzi is currently testing other ALS-associated TDP-43 mutations to see if they affect the helical interactions. He is also working out the protein’s structure inside the granules.

“This is a very nice paper,” wrote Clifford Brangwynne, Princeton University, New Jersey, to Alzforum. “The biophysical relationship between liquid-liquid phase separation and aggregation still remains largely unclear for TDP-43 and other disease-related proteins, but this study provides a much-needed molecular view of the protein-protein contacts that underlie these processes.”

Taylor agreed. “These authors provide novel detail into the molecular basis of phase separation,” he wrote. But he said the question of how the mutations drive disease still remains. It could be by impairing the assembly and function of membraneless organelles such as stress granules, by enhancing the formation of toxic fibrils, or both.

Fawzi cautioned that because the experiments were done in a test tube with pure TDP-43 fragments and because RNA granules in vivo contain different types of protein, it remains to be seen how these interactions play out in a cell. However, he pointed to a recent, complementary finding that was published in the August 2 Cell Reports. It detailed the importance of the same subdomain of TDP-43’s C-terminal end in phase-separated droplets of TDP-43 within HEK293T cells. Broder Schmidt and Rajat Rohatgi at the Stanford School of Medicine in California found that if they deleted the region entirely, the protein could not form phase-separated droplets in the cells. Replacing the conserved helix with a non-helical segment similar to the rest of the disordered domain made it assemble into stable filaments that impaired cell growth. In addition, the ALS-associated variants M337V, N345K, and A382T, all of which somehow change the secondary structure of the conserved region, impaired the formation or function of RNA granules in some way. They also concluded that this conserved helix enables TDP-43 to enter or form these membraneless organelles.

Scientists previously proposed that the TDP-43 helix interacts with RNA-binding domains of other proteins, or that it tethers TDP-43 to the cell membrane, Song wrote to Alzforum. “It will be interesting to figure out the interplay of these roles and their physiological and pathological consequences,” he added. It also will be important to know whether the process of TDP-43 assembly that Fawzi describes is a separate pathway from formation of amyloid fibers, or a different stage of the same pathway. To parse this out, Song proposed designing mutants with stabilized helices and testing their toxicity in neurons.—Gwyneth Dickey Zakaib


  1. The paper is very interesting. As compared to our previous paper (Lim et al., 2016), two points are particularly interesting:

    1.     In 2005, we discovered that previously-claimed “insoluble proteins,” including the most hydrophobic integral membrane fragment in nature, could all be solubilized in pure water (unsalted) (Song, 2009). Now our discovery has been confirmed by many labs worldwide and, in particular, a consortium of Japanese scientists showed that almost all human proteins, including membrane proteins, could be dissolved in pure water for preparing cancer vaccines (Futami et al., 2014). With this discovery, we have successfully studied several previously thought “insoluble” ALS-causing proteins/mutants including TDP-43 N-domain (Qin et al., 2014), prion-like domain (Lim et al., 2016), and SOD1 (Lim et al., 2015; Lim and Song, 2016). 

    Very interestingly, Professor Fawzi’s group used an organic buffer MES (N-morpholinoethanesulfonic acid) of very low ionic strength to mimic pure water to successfully study TDP-43 prion-like domain in their current paper, as they previously did for the FUS prion-like domain (Burke et al., 2015). Our latest study indicates that MES buffer is indeed similar to unsalted water (Lu et al., 2016). This is a very interesting result that most likely can be generally applied to studying other “insoluble proteins” causing ALS or other diseases/aging.

    2.     The most novel discovery is that the helical region is critical to maintaining the dynamic assembly of the wild-type TDP-43 prion-like domain necessary for its functions, while the destabilization of the helix will lead to pathological aggregation. It is very important to note that the helical region is a typical hydrophobic fragment that bears no prion-like sequence, but plays a central role in forming classic amyloid fibers. Our latest results indeed show that a shorter TDP-43 C-terminal domain lacking this hydrophobic region is no longer able to form classic amyloid fibers (Lu et al., 2016). Therefore, Professor Fawzi’s current results, together with our previous (Lim et al., 2016) and latest findings (Burke et al., 2015), reveal a significant difference between the molecular mechanisms for the assembly of the WT and ALS-causing mutants of TDP-43 C-domain: Although both of them form fibrillar structures, as we found (Lim et al., 2016), in the dynamic WT assembly the helical region continues to be a helix, while in the ALS-causing mutants, the helical region transforms into amyloid structures rich in cross-β structures as judged by CD and fluorescence spectra in our previous paper (Lim et al., 2016). Furthermore, as the TDP-43 C-terminal domain contains this hydrophobic helical region and prion-like sequences, it appears that two mechanisms are simultaneously operating in the assembly/amyloid formation of the TDP-43 C-terminal domain: The hydrophobic region follows the classic mechanism for protein-protein interactions/amyloid formation, while the prion-like sequences assemble into a dynamic/reversible hydrogel with a pH-dependent transformation from the intramolecular to intermolecular backbone-sidechain hydrogen networks, as we first proposed (Lim et al., 2016) and now further enforced (Lu et al., 2016). The delineation of the different mechanisms may have important implications in understanding the molecular basis for ALS and other protein-aggregation-causing diseases/aging.

    Furthermore, I also have some thoughts for further investigations:

    1.     The helical region has now been revealed to be potentially involved in three roles: 1) mediation of the dynamic assembly of TDP-43 C-domain as shown in Professor Fawzi’s current paper; 2) interaction with the RNA-binding domains in the context of the full-length TDP-43 (Wei et al., 2016); 3) potential interaction with membranes (Lim et al., 2016). It will be extremely interesting to figure out how these roles interplay and are regulated in cells, and what their physiological and pathological consequences might be. Also, whether the potential to interact with membranes is physiological or just a pathological aberrance needs to be addressed. Recently we found documented a biophysical mechanism (Qin et al., 2013; Lim et al., 2015) whereby ALS-causing mutants of cytosolic SOD1, whose native functions have nothing to do with membranes, can, unbelievably, acquire the capacity to insert into ER membranes to initiate ALS without any detectable protein aggregation (Sun et al., 2015). We have even found that although the native functions of the E. coli protein S1 have nothing to do with membranes, cutting it to mimic the oxidation-induced fragmentation seen during aging unlocks fragments with high amphiphilicity/hydrophobicity that have characteristics of “insolubility” in buffers but suddenly acquire the novel capacity to interact with membranes (Lim et al., 2016). 

    2.     It is also very interestingly to investigate whether the dynamic assembly mediated by the helix formation and pathological formation of amyloid fibers represent two different pathways, or are different stages of the same pathway. It may be possible to clarify this by assessing if mutants designed to significantly stabilize the helix can reduce or completely remove neurotoxicity in vivo.


    . ALS-Causing Mutations Significantly Perturb the Self-Assembly and Interaction with Nucleic Acid of the Intrinsically Disordered Prion-Like Domain of TDP-43. PLoS Biol. 2016 Jan;14(1):e1002338. Epub 2016 Jan 6 PubMed.

    . Insight into "insoluble proteins" with pure water. FEBS Lett. 2009 Mar 18;583(6):953-9. Epub 2009 Feb 20 PubMed.

    . Denatured mammalian protein mixtures exhibit unusually high solubility in nucleic acid-free pure water. PLoS One. 2014;9(11):e113295. Epub 2014 Nov 18 PubMed.

    . TDP-43 N terminus encodes a novel ubiquitin-like fold and its unfolded form in equilibrium that can be shifted by binding to ssDNA. Proc Natl Acad Sci U S A. 2014 Dec 30;111(52):18619-24. Epub 2014 Dec 12 PubMed.

    . Resolving the paradox for protein aggregation diseases: a common mechanism for aggregated proteins to initially attack membranes without needing aggregates. F1000Res. 2013;2:221. Epub 2013 Oct 21 PubMed.

    . Mechanism for transforming cytosolic SOD1 into integral membrane proteins of organelles by ALS-causing mutations. Biochim Biophys Acta. 2015 Jan;1848(1 Pt A):1-7. Epub 2014 Oct 12 PubMed.

    . SALS-linked WT-SOD1 adopts a highly similar helical conformation as FALS-causing L126Z-SOD1 in a membrane environment. Biochim Biophys Acta. 2016 Sep;1858(9):2223-30. Epub 2016 Jul 1 PubMed.

    . 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.

    . Mechanisms of self-assembly and fibrillization of the prion-like domains. bioRχiv. 2016 Jul 25.

    . Inter-domain interactions of TDP-43 as decoded by NMR. Biochem Biophys Res Commun. 2016 Apr 29;473(2):614-9. Epub 2016 Apr 1 PubMed.

    . Translational profiling identifies a cascade of damage initiated in motor neurons and spreading to glia in mutant SOD1-mediated ALS. Proc Natl Acad Sci U S A. 2015 Dec 15;112(50):E6993-7002. Epub 2015 Nov 30 PubMed.

    . Unlocked capacity of proteins to attack membranes characteristic of aggregation: the evil for diseases and aging from Pandora's box. bioRχiv. 2016 Aug 24.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Do Membraneless Organelles Host Fibril Nucleation?
  2. C9ORF72 Mice Point to Gain of Toxic Function in ALS, FTD
  3. Artificial Chaperone Keeps Amyloid-Forming Protein in Check
  4. Form and Function: What Makes α-Synuclein Toxic?

Paper Citations

  1. . Phase separation by low complexity domains promotes stress granule assembly and drives pathological fibrillization. Cell. 2015 Sep 24;163(1):123-33. PubMed.
  2. . 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.
  3. . ALS-Causing Mutations Significantly Perturb the Self-Assembly and Interaction with Nucleic Acid of the Intrinsically Disordered Prion-Like Domain of TDP-43. PLoS Biol. 2016 Jan;14(1):e1002338. Epub 2016 Jan 6 PubMed.

Further Reading


  1. . Inter-domain interactions of TDP-43 as decoded by NMR. Biochem Biophys Res Commun. 2016 Apr 29;473(2):614-9. Epub 2016 Apr 1 PubMed.
  2. . Structural insights into the multi-determinant aggregation of TDP-43 in motor neuron-like cells. Neurobiol Dis. 2016 Oct;94:63-72. Epub 2016 Jun 16 PubMed.
  3. . An insoluble frontotemporal lobar degeneration-associated TDP-43 C-terminal fragment causes neurodegeneration and hippocampus pathology in transgenic mice. Hum Mol Genet. 2015 Dec 20;24(25):7241-54. Epub 2015 Oct 16 PubMed.

Primary Papers

  1. . 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.
  2. . In Vivo Formation of Vacuolated Multi-phase Compartments Lacking Membranes. Cell Rep. 2016 Aug 2;16(5):1228-36. Epub 2016 Jul 21 PubMed.