From corticobasal degeneration to frontotemporal dementia, tauopathies come in many guises. How disruptions in a single protein cause such varied diseases remains a mystery. Results from a new study support the idea that unique conformations of misfolded tau may be to blame. Researchers led by Marc Diamond of the University of Texas Southwestern Medical Center in Dallas report that tau strains from various human and mouse tissues spread at different rates in different areas of the mouse brain, with some settling in specific regions, such as the hippocampus. Published in the November 23 Neuron, the research builds on the idea that the temporal and spatial spread of pathology in different tauopathies might be explained by specific tau strains.

“This is a very exciting paper. It reveals even more intriguing insights into the complexity of tau that could well extend to other disease-associated proteins that appear to propagate,” said Karen Duff of Columbia University in an email.

Tau is a microtubule-associated protein thought to become toxic when misfolded forms aggregate, driving neural and glial dysfunction. Diamond has argued that tau behaves like prions, stably passing on unique conformations during cell replication and “corrupting” fellow proteins upon contact (see Aug 2013 conference news). Previously, his group reported that several such “strains,” propagated faithfully through cell lines, led to distinctive aggregate inclusions in cultured cells, and drove distinguishing cellular pathology in a transgenic mouse model (see May 2014 news). 

Strains Go Their Separate Ways. Over time, tau strains DS1 to DS10 spread differently from their common injection site in hippocampus. [Courtesy of Marc Diamond.]

For this study, co-first author David Sanders wanted to examine a broader set of tau strains to assess their potential for influencing regional pathology and speed of progression as is seen in different tauopathies. Sanders dipped into a library of tau strains he had previously created by exposing HEK293 cells expressing P301S tau to tau aggregates from various sources, including mouse and human brain samples. He evaluated 90 cell lines that stably propagated tau aggregates, paring them down to 18 seemingly distinct strains based on their morphologies, seeding rates, and resistance to proteolysis. Each strain maintained its specific characteristics through multiple passages. 

“What is especially remarkable is the fidelity whereby aggregates continue to amplify themselves,” said Sanders. He noted that these 18 strains likely represent a fraction of all possible tau strains, but also acknowledged that he cannot yet rule out whether some of the more similar strains may be identical.

After Sanders characterized the 18 strains in vitro, co-first author Sarah Kaufman analyzed how they propagate in the brain of PS19 transgenic mice, which express full-length human mutant tau carrying the P301S mutation. Kaufman injected each strain into the hippocampi of two- to three- month-old mice, and eight weeks later examined tau pathology there. She found stark differences. While certain strains seeded neurofibrillary tangle-like aggregates in the CA1 and CA3 regions, some strains yielded wisp-like tau aggregates, others “dotted” the mossy fibers alone, and still others led to multiple forms of tau throughout the hippocampus.

Next, Kaufman looked at how tau spread across other brain regions by injecting strains into six different sites: the visual cortex, thalamus, sensory cortex, caudate/putamen, inferior colliculus, and hippocampus. After five weeks, only three of six strains tested drove tau pathology in all regions. One caused pathology in every region except the inferior colliculus. Another seeded aggregation in the caudate/putamen, hippocampus, and thalamus, and the last only in the hippocampus. The extent of pathology in each region also varied among strains. The scientists used blinded assessors who rated AT8 staining as none, low, medium, or high.

In further analyses Kaufman and colleagues examined the speed of propagation. After injecting select strains into the hippocampus, they tracked their spread to nearby regions by examining tau pathology four, eight, and 12 weeks later. While certain strains spread widely and quickly, even when the tau inoculation was diluted, others took hold more slowly and only in specific areas, with one strain being restricted to mossy fibers (see image above).

“It’s amazing that just by changing the strain type, you can basically create completely different tauopathy syndromes in the mice,” said Diamond. The findings support the concept that a particular strain may drive a particular disease or syndrome; however, the scientists did not attempt to link the pathologies they observed to specific, real-life human tauopathies, Diamond said. The mice underwent carefully controlled procedures that would be very different from what a person with a tauopathy experienced, for example injections of single strains, whereas human tauopathies may well result from multiple strains, Diamond noted.

Nevertheless, Diamond added that determining the molecular structure of these different strains could help direct drug design and the creation of PET ligands specific to known human tauopathies. “This paper is saying that one size does not fit all,” he said. “You might need to make conformation-specific therapies or diagnostics.”

The authors acknowledged that the need to use tau transgenic mice to get tau strains to seed was a limitation. Lary Walker of Emory University, Atlanta, noted there is a strong case for using those models, but added that it makes extrapolating the results to sporadic human tauopathies, such as Alzheimer's disease, a bigger jump.

Even so, Walker considers the evidence for the study’s main point compelling, namely that differing tauopathies are likely driven by different strains. He also agreed with Diamond’s claim that more than one strain might be needed to cause human disease. “Even a tauopathy that is diagnosed as one disorder, may in fact involve multiple strains of misfolded tau that can influence the course of the disease in different ways,” said Walker. “This model is a way of teasing addressing that and figuring out exactly what's going on.” —Lindzi Wessel

Lindzi Wessel is a writer based in San Jose, California.



  1. This is a very exciting paper that reveals even more intriguing insights into the complexity of tau that could well extend to other disease-associated proteins that appear to propagate. The paper expands on the Diamond lab's previous ideas and findings about tau that really start to explain why even the related tauopathies (for example, the 4R tauopathies) show such phenotypic diversity, with clinical symptoms ranging from social/emotional to memory to motoric. Because these studies utilize an in vitro step to generate the tau strains, they do not aim to reproduce the biology of tau forms that we expect exist in the intact brain. The strains should be considered instead as carriers of information that makes them behave differently in vivo and in vitro, which will allow us to address fundamental questions about phenotypic diversity of the tauopathies. To address human relevance, important follow-up studies would include looking more at aspects of cellular and regional vulnerability, functional consequences, and impact on cognitive measures.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Are Protein Strains The Cause of Different Tauopathies?
  2. Like Prions, Tau Strains Are True to Form

Research Models Citations

  1. Tau P301S (Line PS19)

Further Reading

No Available Further Reading

Primary Papers

  1. . Tau Prion Strains Dictate Patterns of Cell Pathology, Progression Rate, and Regional Vulnerability In Vivo. Neuron. 2016 Nov 23;92(4):796-812. Epub 2016 Oct 27 PubMed.