Extracts from postmortem AD brains or old AD transgenic mice can seed Aβ deposition in mice that would typically rack up amyloid much later in life (ARF related news story on Meyer-Luehmann et al., 2006), and even in mouse models that normally would not develop plaques (ARF related news story on Morales et al., 2011). More recently, researchers demonstrated that a pyroglutamylated form of Aβ can transmit its neurotoxicity to normal Aβ42 peptides in mouse neuronal cultures (ARF related news story on Nussbaum et al., 2012). But brain extracts contain much more than Aβ, and none of many defined synthetic preparations tested to date has seeded amyloid deposition. Can an inoculation of pure Aβ drive its aggregation in vivo?

Lead authors Jan Stöhr and Joel Watts, and colleagues report that both purified and synthetic Aβ are sufficient to seed Aβ aggregation in vivo. First, the researchers purified Aβ aggregates from the brains of old APP23 or TgCRND8 AD mice, and injected them into the brains of a mouse model that allows in-vivo monitoring of amyloid accumulation. Watts made this strain by crossing APP23 transgenics with animals that express a luciferase (luc) reporter under the control of the glial fibrillary acidic protein (GFAP) promoter (ARF related news story on Watts et al., 2011). In APP23/GFAP-luc mice, “glowing” glia ramp up their signal around 14 months (~416 days) of age, illuminating brain areas with amyloid buildup and AD-like neuropathology.

In APP23/GFAP-luc mice inoculated at two months of age with purified brain Aβ, the bioluminescence signal showed up at 7.7 months (about 170 days after injection). By comparison, APP23/GFAP-luc mice receiving crude APP23 or TgCRND8 brain extract showed no signs of amyloidosis until 10.3 months (~250 days after injection). However, though the purified Aβ prep was much more effective at seeding aggregation, “it could be argued that something might be co-purifying with the Aβ,” Giles told Alzforum.

To prove that Aβ alone seeds amyloidosis, the researchers injected APP23/GFAP-luc mice with synthetic forms of the peptide. They used either wild-type Aβ1-40, or an S26C mutant forming dimers that block long-term potentiation (Shankar et al., 2008) and form neurotoxic protofibrils (O’Nuallain et al., 2010). The synthetic preps induced amyloid deposition about 230 days after injection. This makes them much less potent than the crude brain extract or the Aβ purified from it, since those preparations contained 100 and 15 times less Aβ, respectively, than the synthetic inoculations did. But the main point is that synthetic Aβ did it at all.

The study is “extremely nice,” said Mathias Jucker of the University of Tübingen, Germany. “It really shows that Aβ is the amyloid-inducing agent. In our studies we assumed it was, but never had 100 percent proof.” Jucker and others have done similar experiments with Aβ-containing brain extracts but did not yet demonstrate that seeding works with synthetic Aβ. Jucker said the current study succeeded because the authors used more material and waited longer. Claudio Soto of the University of Texas Medical School, Houston, agreed that the main difference was quantity—they injected “about 50 times more (synthetic Aβ) than we did.”

Giles and colleagues contend that the new findings “provide incontrovertible evidence that Aβ aggregates are prions and that formation of Aβ prions does not require additional proteins or cofactors.” In a commentary published June 22 in Science, Prusiner highlights recent work suggesting that a diverse group of proteins form prions in neurodegenerative diseases.

Some scientists question if “prion” is a suitable term for Aβ or other amyloid-forming proteins such as tau and α-synuclein, which exhibit seeding behavior in cell culture and animal models (see Clavaguera et al., 2009; Liu et al., 2012; de Calignon et al., 2012; Luk et al., 2012; Desplats et al., 2009). The squabble may center more around semantics than science. Though these proteins can self-propagate across cells and tissues, “we advise caution in using the words ‘prion’ or ‘prionoid’ or ‘prion-like,’” wrote John Hardy and Tamas Revesz of University College London, U.K., in a commentary for the May 31 New England Journal of Medicine (Hardy and Revesz, 2012). “Prion disease can spread from animal to animal, from animal to human, and from human to human. We also know that the prion infectious agent is stable in the environment for many years. We have no evidence that any of this is true for Alzheimer’s disease, Parkinson’s disease, or frontotemporal dementia and their etiologic agents, and we should avoid the implications that the word ‘prion’ carries.”

Giles finds the fuss over terminology overblown. “It’s just a word—like ‘virus.’ Just because you use the word virus, it doesn’t mean ‘Ebola virus.’ Prions are proteins that can change conformation and self-propagate that altered conformation. That’s how the term has been used in the scientific community for 20 years,” he told Alzforum.

Jucker agrees, though with a caveat. “In a closed room of structural biologists, molecular biologists, and people in the amyloid field, I think it’s perfectly fine to call Aβ a prion because we all know what it means,” said Jucker. “But if you’re with people who think about infectivity and what it means for human beings, then I think we should be very careful.”

Others say the key question is not so much if Aβ, tau, and α-synuclein behave like prions, but what makes prions behave so differently. “If you put amyloid coming from the brain of a Creutzfeldt-Jakob disease patient into a non-transgenic mouse containing endogenous prion protein, it’s going to propagate very fast and kill the mouse,” said Eliezer Masliah of the University of California, San Diego. “But if you inject the Aβ or synuclein prion, you’re not going to see much—unless you put it into a mouse that is susceptible, like a young transgenic mouse.”

The question of what causes susceptibility to prion infection was addressed in another paper published this week in PNAS. Surachai Supattapone and colleagues at Dartmouth Medical School, Hanover, New Hampshire, found that cofactor molecules are critical for infectivity, conformation, and other key properties of prions. Because the Aβ aggregates in the UCSF study required “special circumstances such as overexpression of the amyloidogenic protein (in the host) and high concentrations of inoculum to induce the spread of their misfolded state … I would say no, they are probably not prions,” Supattapone wrote in an e-mail to Alzforum.

Furthermore, although the induction of Aβ aggregation “is clearly possible in experimental animals, the main question for the future is whether this happens in real life in humans,” Soto said. “The current study does not address this issue.” Soto and colleagues have a paper under review exploring whether Aβ aggregation can be transmitted through blood transfusion and other real-life conditions. For their part, the authors are trying to find specific Aβ conformations that are most efficient at seeding aggregation.—Esther Landhuis.

References:
Stöhr J, Watts JC, Mensinger ZL, Oehler A, Grillo SK, DeArmond SJ, Prusiner SB, Giles K. Purified and synthetic Alzheimer’s amyloid β (Aβ) prions. PNAS Early Edition. 2012 June 18. Abstract

Deleault NR, Walsh DJ, Piro JR, Wang F, Wang X, Ma J, Rees JR, Supattapone S. Cofactor molecules maintain infectious conformation and restrict strain properties in purified prions. PNAS Early Edition. 2012 June 18. Abstract

Prusiner SB. A Unifying Role for Prions in Neurodegenerative Disease. Science. 2012 June 22;336:1511-1513. Abstract