In two papers in the June 30 Proceedings of the National Academy of Sciences, researchers led by Stanley Prusiner at the University of California, San Francisco, add to the growing body of evidence suggesting that Aβ curls into distinct shapes, or strains, that propagate through the brain by templated protein misfolding.

In the first paper, Prusiner and colleagues report that different polymerization conditions cause synthetic Aβ42 to assume distinct forms, which produce different types of plaques when inoculated into mouse brain. The second paper compares Aβ aggregates isolated from Alzheimer’s brains. In human brain, mutant Aβ folds differently than wild-type, and causes a distinct pathology when seeded into transgenic mouse brain, the authors found. Moreover, mutant human Aβ aggregates transfer their properties to wild-type human Aβ in these animals. This corrupted wild-type Aβ can then seed the same pathology in new mice, suggesting that the aggregates maintain and pass on a specific structure.

“Key findings of these studies are that it is possible to generate biologically active strains of aggregated Aβ from synthetic peptides, and that both synthetic and AD-derived Aβ strains can be seeded and propagated in vivo,” wrote Lary Walker at Emory University, Atlanta. Walker was not involved in the research. He noted that the findings have important implications. “If disease-relevant, human-like Aβ strains can be generated in mouse models, the resulting mice might more effectively predict the efficacy of therapeutics and diagnostics in humans,” he suggested (see full comment below).

Synthetic Aβ42 injected into mouse brain seeds small, compact plaques that contain mostly Aβ42 (green, left panel); when polymerized with the detergent SDS, the plaques become large, diffuse, and contain more Aβ40 (red, middle panel); this matches the appearance of plaques seeded by synthetic Aβ40 polymerized with SDS (right panel). Image courtesy of Stöhr et al., 2014, PNAS.

Other pathogenic proteins, including tau and α-synuclein, are now believed to spread through the brain by templated misfolding (see Nov 2012 news storyNov 2012 conference storyAug 2013 conference storyMay 2014 news story). Aβ also propagates along synaptic pathways from initial seeds (see Jul 2009 news story; Nov 2010 news story). Previously, Prusiner and colleagues demonstrated that synthetic or purified Aβ could kick off amyloid deposition in AD model mice (see Jun 2012 news story).

Does Aβ form distinct strains, however? Some prior evidence suggests as much. Normal Aβ appears to be corrupted by pyroglutamate Aβ, an N-terminally modified form that is particularly sticky and aggregation-prone (see May 2012 news story). In addition, Robert Tycko and colleagues at the National Institutes of Health, Bethesda, Maryland, found that Aβ from two AD patients seeded aggregates with distinct structures in vitro, suggesting that each person had harbored a different strain (see Sep 2013 news story).

To investigate the question of strains further, Prusiner and colleagues turned to synthetic Aβ and experimented with different polymerization conditions. When first author Jan Stöhr exposed Aβ42 monomers to a sodium phosphate solution, the peptide assembled into numerous short fibrils. When he added a splash of the detergent SDS to the polymerization solution, Aβ42 exclusively formed long fibrils. This implied to the authors that the two conditions might be giving rise to different strains of aggregated peptide. Synthetic Aβ40 aggregated into long fibrils under either condition.

To see how these strains would behave in vivo, the authors injected the aggregates into six-week-old transgenic APP23 mice, which overproduce wild-type human Aβ. Eleven months after inoculation, an age when control TgAPP23 animals have little brain pathology, inoculated mice displayed extensive Aβ deposits, but their nature varied greatly depending on which inoculate the mice had received. Animals injected with short Aβ42 fibrils formed in the absence of SDS had numerous small, compact plaques loosely scattered along the corpus callosum next to the CA1 region of the hippocampus. These plaques contained predominantly Aβ42 (see image above). Animals injected with long Aβ42 fibrils formed in the presence of SDS, or with Aβ40 fibrils, had fewer, larger plaques clustered along the corpus callosum. They contained primarily Aβ40 and had a diffuse structure (see image above). The latter mice also developed less astrocytic gliosis than mice inoculated with short Aβ42 fibrils. In ongoing work, the authors are testing whether these characteristic pathologies repeat when deposits are propagated in new mice. That would indicate that the strains stay stable. 

In the second paper, first author Joel Watts and colleagues describe strains isolated from human brain. The authors obtained homogenates from postmortem brains of two patients with sporadic Alzheimer’s, one person with the Arctic APP mutation (which changes the sequence of Aβ), and one person with the Swedish APP mutation, which results in overproduction of wild-type Aβ. When injected into APP23 mice, the Arctic samples produced distinct disease characteristics. Aβ deposits formed more quickly and contained more Aβ42 than those induced by seeds from the other patients. Arctic seeding material also induced more amyloid in blood vessels. This cerebral amyloid angiopathy (CAA) contained high levels of Aβ38, and had a “furry” appearance, with deposits extending out from the vessels. AD patients with the Arctic mutation also have extensive CAA that includes Aβ38 (see Moro et al., 2012). 

Notably, because APP23 mice produce only wild-type Aβ, these findings suggest that the Arctic Aβ transferred its structure to the normal peptide. Homogenates from the brains of these Arctic-seeded mice induced the same disease phenotype when injected into young APP23 animals, indicating that this Aβ conformation maintains itself through serial passaging. Brain homogenate from one of the sporadic AD patients produced a mix of “furry” and normal CAA, implying that the Arctic structure may arise spontaneously in people who do not carry the Arctic mutation, suggest the authors.

The propagation of a distinct pathology from one animal to another provides evidence of strains, agreed Marc Diamond at Washington University in St. Louis. To be certain, aggregates from the different generations of mice should be purified and structurally analyzed to show that they maintain a stable conformation, Diamond suggested.—Madolyn Bowman Rogers


  1. The prion paradigm has considerable explanatory power for the ontogeny of human neurodegenerative diseases, which mostly appear to emerge from the endogenous misfolding and seeded aggregation of disease-specific proteins. These two papers add to burgeoning evidence that disease-associated proteins can assume subtly different three-dimensional shapes, i.e. "strains," and that shape influences the pathogenic features of the resulting aggregates. Key findings of these studies are that it is possible to generate biologically active strains of aggregated Aβ from synthetic peptides, and that both synthetic and AD-derived abeta strains can be seeded and propagated in vivo. If strains are differentially sensitive to therapeutics, developing a universally effective intervention could be difficult. On the other hand, if disease-relevant, human-like Aβ strains can be generated in mouse models, the resulting mice might more effectively predict the efficacy of therapeutics and diagnostics in humans.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Toxic Synuclein Corrupts Native in Wild-Type Mice
  2. Do Tau "Prions" Lead the Way From Concussions to Progression?
  3. Are Protein Strains The Cause of Different Tauopathies?
  4. Like Prions, Tau Strains Are True to Form
  5. Aβ the Bad Apple? Seeding and Propagating Amyloidosis
  6. Insidious Spread of Aβ: More Support for Synaptic Transmission
  7. Aβ Sufficient for Seeding—But Is It a Prion?
  8. Pyromania: PyroGluAβ Toxicity is Prion-Like, Depends on Tau
  9. Does Aβ Come In Strains? Glimpse Into Human Brain Suggests Yes

Research Models Citations

  1. APP23

Mutations Citations

  1. APP E693G (Arctic)
  2. APP KM670/671NL (Swedish)

Paper Citations

  1. . APP mutations in the Aβ coding region are associated with abundant cerebral deposition of Aβ38. Acta Neuropathol. 2012 Dec;124(6):809-21. PubMed.

Further Reading

Primary Papers

  1. . Distinct synthetic Aβ prion strains producing different amyloid deposits in bigenic mice. Proc Natl Acad Sci U S A. 2014 Jul 15;111(28):10329-34. Epub 2014 Jun 30 PubMed.
  2. . Serial propagation of distinct strains of Aβ prions from Alzheimer's disease patients. Proc Natl Acad Sci U S A. 2014 Jul 15;111(28):10323-8. Epub 2014 Jun 30 PubMed.