For instant microtubules, just add a droplet of … tau. In the third recent paper to report that tau undergoes liquid-liquid phase separation, researchers describe how tau droplets subsume tubulin and—shazam!—microtubules grow and squelch the droplets from round to oblong. The researchers, led by Stefan Diez and Anthony Hyman, both of the Max Planck Institute of Molecular Cell Biology and Genetics in Dresden, Germany, propose that tau droplets seed microtubule growth throughout the cell.

  • Tau droplets recruit tubulin.
  • Droplet tubulin forms microtubules.
  • This explains how microtubules sprout in different cellular locales. 

Combined with other recent reports, the findings highlight a pivotal role of liquid-liquid phase separation in tau function and pathology, commented Markus Zweckstetter of the German Center for Neurodegenerative Diseases in Göttingen, senior author on one of the other recent tau droplet papers.

“These are exciting times in tau biology,” Zweckstetter said.

While others universally praised the beauty of this in vitro study, they also questioned if the phenomena it captures happen in vivo. “I worry that the findings reflect events that occur in a test tube and cell culture, but may not reflect what happens in ‘real life,’” commented John Trojanowski of the University of Pennsylvania in Philadelphia.

Nope, Not Shooting Stars. Added to tau droplets, tubulin polymerizes into microtubules, wrenching drops into liquid strings. [Courtesy of Hernandez-Vega et al., Cell Reports, 2017.]

Tau is an intrinsically disordered protein. Two recent studies indicated that it undergoes liquid-liquid phase separation (LLPS) in vitro. The protein reportedly condenses into droplets via electrostatic interactions with negatively charged molecules, including RNA, heparin, or even via phosphate groups on tau itself (Jul 2017 news and Aug 2017 news). Scientists believe that LLPS of other disordered proteins gives structure to membraneless organelles, such as stress granules and the nucleolus, that play essential roles in cells. Might tau droplets be equally important for cell physiology? 

First author Amayra Hernández-Vega and colleagues asked if tau LLPS would influence the protein’s ability to stabilize microtubules. These struts form when tubulin polymerizes, a phenomenon best characterized at centrosomes, where tubulin gathers during mitosis. Interestingly, some evidence suggests that centrosomes also form membraneless liquid droplets to concentrate tubulin and other proteins (Zwicker et al., 2014). Being post-mitotic, neurons lack centrosomes; even so, they grow dense microtubule tracts, especially in axons, to support the transport of cargo over long distances. The researchers hypothesized that tau droplets could focus tubulin in these locales much like centrosomes do, buttressing the vast microtubule network.

First, the researchers tested if tau could form droplets. Using a purified human tau isoform containing four microtubule repeat domains, they found, much as others have, that using molecular crowding agents such as dextran, polyethylene glycol, or Ficoll to mimic the viscosity inside of a cell forced tau into liquid droplets. Salt stymied this process, suggesting that electrostatic interactions held the droplets together.

The researchers next added tubulin into the mix. In the presence of dextran, α/β dimers of tubulin partitioned into tau droplets, where it concentrated more than 10-fold. The researchers hypothesized that electrostatic interactions between tubulin, a negatively charged protein, and tau, which is positively charged, facilitated droplet formation. Tubulin also promoted formation of further droplets.

Would these droplets affect tubulin polymerization or the growth of microtubules? To find out, the researchers mixed together guanidine triphosphate (GTP), which is required for tubulin polymerization, tubulin, and tau droplets. Within minutes, tubulin melded into the droplets then polymerized, stretching the round droplets into long skinny ones (see video above). Seen through a microscope, many microtubules bundled together, apparently held by a coating of tau. Photo-bleaching experiments revealed that while microtubules remained trapped within their liquid cages, tau roamed freely in and out of the elongated droplets. Regardless of how much crowding agent the researchers added to the mix, microtubules never formed unless tau was present.

To test if condensation in the droplets was the sole driver of microtubule formation, the researchers mixed tau and tubulin at concentrations seen in droplets, and left out crowding agents. Short microtubules formed, but did not elongate or bundle together. The researchers concluded that something special about droplets, beyond mere concentration, promoted microtubule growth. In fact, when added to existing microtubules, tau/tubulin/GTP droplets rapidly bound and sprouted new polymers alongside the existing ones. To confirm that tau’s presence in the droplets was required to keep these microtubule networks stable, they added heparin—a polyanion known to interfere with tau-tubulin interactions and promote tau fibrillization—to microtubule-laden strands of liquid tau. This not only severed ties between tau and tubulin, but unbundled and depolymerized the microtubules. Once separated from tubulin, tau reformed spherical droplets on its own (see video below).

Tubules Interrupted. Heparin instantly disrupts tau (green/yellow) from microtubules (red). The tau condenses into droplets, while the microtubules eventually fall apart.

Diez told Alzforum that the exact nature of tau and tubulin’s interaction within the droplets is still unclear. However, he proposed that individual tubulin molecules may bind tau via their negatively charged “E-hook domain.” This electrostatic interaction then condenses tubulin to a critical concentration that allows it to polymerize into microtubules. On mature microtubules, the E-hooks project outward like hairs, Diez explained, which might serve to further attract tau. 

Droplets to Bundles. In the proposed model, tau condenses into droplets, then tubulin joins. There, tubulin polymerizes into microtubules, which form bundles and deform the droplets. [Courtesy of Hernandez-Vega et al., Cell Reports 2017.]

Diez added that phase separation is a finely tuned process, meaning that it could be modulated by any number of other molecules, such as RNA, or post-translational modifications, such as phosphorylation. Both of these have been demonstrated in recent studies. What happens inside neurons is likely to be more complicated, he said. He speculated that the local concentration of tau droplets in neurons could dictate where new microtubules form.

“Hernandez-Vega et al. convincingly demonstrate that the role of tau droplets is not just that of a protein concentrator, but that of actively promoting elongation and growth of tubulin polymerization,” wrote Kenneth Kosik and Songi Han of the University of California, Santa Barbara in an email to Alzforum. “The mechanism for the latter remains to be understood, but is generally assumed to be due to the effects of confinement and diffusion-limited reaction processes.” Kosik and Han, who co-led another recent study describing tau droplets, commented that tau droplets in promoting microtubule growth does not contradict the idea they and others have proposed—that tau sequestration in droplets is initially beneficial, but could also become an intermediate step to tau aggregation.

Zweckstetter was intrigued that tubulin—much like RNA or heparin—promoted the formation of tau droplets. He wondered how the tubulin droplet environment might change tau’s conformation. His lab recently reported that short stretches of tau’s microtubule binding domains form a hairpin-like structure upon interaction with tubulin. “Droplet formation might thus favor this hairpin formation and promote cross-linking of microtubule filaments into bundles,” he wrote.

Benjamin Wolozin of Boston University called the study provocative, but noted that tau blanketed microtubules in the liquid phase, whereas tau proteins dot microtubules in an evenly spaced manner in vivo. “From a biophysical perspective, the results are fascinating, but it is incumbent on the researchers to discriminate between a test-tube phenomenon and something truly biological,” he said.—Jessica Shugart





  1. Liquid-liquid phase separation (LLPS) of intrinsically disordered, low-complexity proteins is a widespread biophysical phenomenon, and recently has been extended to include the microtubule binding protein tau. The most pressing and unresolved questions facing the studies of LLPS is its physiological role in vivo. The study by Hernandez-Vega et al. shows that tau from insect cells, which is highly phosphorylated, is capable of self-coacervating under conditions of molecular crowding. LLPS was enhanced in the presence of tubulin, as well as RNA, both of which are polyanions. Most strikingly, tubulin favorably partitioned into the droplet phase of tau, and microtubule polymerization was promoted within this phase containing high local concentrations of tubulin and tau. Polymerization occurred even when tubulin was added under droplet-forming conditions at a concentration an order of magnitude lower than typically critical for polymerization. Moreover, Hernandez-Vega et al. convincingly demonstrate that the role of tau droplets is not just that of a protein concentrator, but that of actively promoting elongation and growth of tubulin polymerization. The mechanism of the latter remains to be understood, but is generally assumed to be due to the effects of confinement and diffusion-limited reaction processes. The mechanics of such reactions have been modeled by the reaction-diffusion master equation used to study biochemical reactions in living cells (Hellander and Petzold, 2017). 

    The presence of heparin deactivated tubulin polymerization. We believe this was due to competitive binding of the polyanion, heparin, which more favorably binds to tau than does tubulin or RNA.

    The question of the physiological role of tau droplets, and for that matter all LLPS processes, is still an open one. However, this study opens up another possibility, namely that of regulating non-centrosomal cytoskeleton polymerization and microtubule bundling. This is a complementary thesis to that of Ambadipudi et al., who proposed that tau droplet formation is a precursor and intermediate phase toward pathological tau aggregation, and that by Zhang et al., that makes similar observations as in Ambadipudi et al. to propose that tau droplet formation is initially protecting and sequestering tau from aggregation (Ambadipudi et al. 2017Zhang et al., 2017).  

    Songi Han of the University of California Santa Barbara is the co-author of this comment.


    . Reaction rates for reaction-diffusion kinetics on unstructured meshes. J Chem Phys. 2017 Feb 14;146(6):064101. PubMed.

    . Liquid-liquid phase separation of the microtubule-binding repeats of the Alzheimer-related protein Tau. Nat Commun. 2017 Aug 17;8(1):275. PubMed.

    . RNA stores tau reversibly in complex coacervates. PLoS Biol. 2017 Jul;15(7):e2002183. Epub 2017 Jul 6 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Tau Hooks Up with RNA to Form Droplets
  2. More Droplets of Tau

Paper Citations

  1. . Centrosomes are autocatalytic droplets of pericentriolar material organized by centrioles. Proc Natl Acad Sci U S A. 2014 Jul 1;111(26):E2636-45. Epub 2014 Jun 16 PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Local Nucleation of Microtubule Bundles through Tubulin Concentration into a Condensed Tau Phase. Cell Rep. 2017 Sep 5;20(10):2304-2312. PubMed.