Alzheimer's Disease: Return of the Prion Hypothesis


by Jim Schnabel

February 17, 2012

Recent experiments in mice have added weight to the idea that Alzheimer’s is driven by an infection-like spread of protein aggregates in the brain. Although it may seem startling, this hypothesis has been a popular one among Alzheimer’s researchers for the past several years (see “Alzheimer's Protein Shows Prion-like Infectiousness”). In fact, it represents a return to one of the first modern theories about Alzheimer’s: that it is like a weakly-transmissible prion disease.

“There has been a resurgence of this sort of thinking, because there is now real evidence of the potential transmissibility of Alzheimer’s,” says Thomas Wisniewski, a prion and Alzheimer’s researcher at New York University School of Medicine. “In fact, this ability to transmit an abnormal conformation is probably a universal property of amyloid-forming proteins.”

1970s–1980s: Alzheimer’s as a suspected prion disease

In the 1970s and early 1980s, researchers noted basic similarities between Alzheimer’s disease and transmissible spongiform encephalopathies (TSEs) such as scrapie and Creutzfeldt-Jakob disease. Both types of disease leave behind tough, fibrous, protein-containing deposits—called amyloids for the common molecular structures they contain. In Alzheimer’s, researchers knew of two major amyloids: a rounded “amyloid plaque” outside cells and “neurofibrillary tangles” inside cells. Moreover, TSEs create a spongelike pattern of holes in the brain, and this “spongiform” pathology was seen, albeit to a much lesser extent, in the brains of some people with Alzheimer’s.

At the time, many thought that TSEs are caused by “slow viruses.” However, J.S. Griffith, Carlton Gajdusek, Stanley Prusiner and others proposed that TSEs might be “infections” not of viruses or bacteria or other microbes, but of self-replicating structures made only of protein. Gajdusek termed these “infectious amyloids,” but the term “prion,” introduced by Prusiner (also a Dana Alliance member) in 1982, is the one that stuck:

Because the novel properties of the scrapie agent distinguish it from viruses, plasmids, and viroids, a new term “prion” is proposed to denote a small proteinaceous infectious particle which is resistant to inactivation by most procedures that modify nucleic acids [i.e., DNA or RNA]. Knowledge of the scrapie agent structure may have significance for understanding the causes of several degenerative diseases.

TSEs could be transmitted reproducibly, for example by injecting brain matter from a scrapie-infected sheep into the brain of a healthy sheep. But Gajdusek tried similar experiments with Alzheimer’s brain extracts and chimpanzees, and reported in 1980 that there was no reliable transmission of disease.

Then in 1984, George Glenner and C.W. Wong, and in the following year, Colin Masters and Konrad Beyreuther, isolated the Alzheimer’s plaque protein. They noted that it was a different, much smaller protein than the one isolated from scrapie amyloids. As Alzheimer’s and TSE researchers focused on the study of these separate proteins, theories for the two types of diseases began to diverge, and the “prion hypothesis” for Alzheimer’s was largely forgotten.

1980s to 1990s: Plaques cause Alzheimer’s

Masters and Beyreuther made antibodies to the Alzheimer’s plaque protein, which came to be known as amyloid beta (A-beta). With these antibodies they were able to label plaques in autopsied brains much more sensitively than pathologists had been able to do using traditional stains for amyloid. In 1988 they reported that A-beta plaques were common and widespread in the sampled brains of elderly people, and less dense deposits of A-beta were even detectable among some neurologically healthy middle-aged people. Thus it was clear that Alzheimer’s features a slow accumulation of A-beta plaques throughout the brain, with symptoms appearing only in the last decade of the disease.

Beyreuther and colleagues also found that the A-beta protein is a fragment—the remainder after a cutting by enzymes—of a much larger neuronal membrane protein (dubbed “amyloid precursor protein,” or APP), and that APP is coded by a gene on chromosome 21. They knew that people with Down syndrome make excess copies of APP along with other chromosome 21 proteins; have Alzheimer’s-like A-beta plaque deposits in their brains starting in their 30s; and frequently suffer from a form of early-onset dementia. This again hinted that A-beta is the toxic agent that causes Alzheimer’s.

In the early 1990s, new reports seemed to confirm this suspicion. A-beta seemed toxic to neurons in the lab dish. APP mutations were found in families with early-onset Alzheimer’s. And mice that were genetically engineered to overproduce APP developed Alzheimer’s-like brain pathology. Alzheimer’s research now burst into the public consciousness as never before, along with the hope that the disease soon would be cured.

Early 1990s to early 2000s: Confusion and debate

Most of the APP-overexpressing mouse models created in the early 1990s developed plaques, but did not develop the other major Alzheimer’s amyloid, neurofibrillary tangles, which are made of tau protein and appear inside affected neurons. These “Alzheimer’s mice” also failed to show the profound neuronal losses and memory failures seen in human Alzheimer’s patients. The one exception, reported in Nature in late 1991, was retracted in early 1992 amid evidence that it was fraudulent.

Most families with APP mutations were found not to overexpress A-beta (as had been thought initially) but to overexpress a comparatively rare, aggregation-prone version, known as A-beta-42. Initial reports that A-beta is toxic to neurons also were hard to reproduce; later experiments suggested that the protein becomes toxic only when lab-dish conditions allow it to start forming aggregates. Yet the final aggregates in purified plaques seemed at most weakly toxic to neurons—if they were toxic at all.

Autopsy studies in the 1990s also cast doubt on the A-beta hypothesis. They found that most A-beta plaque material appeared in the brain well before the onset of dementia. The tau-based tangles correlated better with dementia, and spread through the brain in a “distribution pattern permitting the differentiation of six stages,” as neuropathologists Heiko and Eva Braak wrote in 1991.

Many researchers still saw the plaques as the earlier drivers of the disease, and so the plaques were the targets for the first experimental treatments aimed at stopping Alzheimer’s. Yet these treatments turned out to work poorly, if at all. Even an active A-beta vaccine that strongly reduced plaques in patients’ brains seemed unable to slow the worsening of dementia.

Were A-beta plaques the wrong target? Were the tau tangles the true culprits? Or were both these amyloids “red herrings”?

Late 1990s to now: The oligomeric prion hypothesis

In early experiments with A-beta, Beyreuther and others had found evidence that the protein, besides forming long, plaque-making fibril aggregates, can cluster into tiny, soluble “oligomers” made of comparatively few copies of A-beta. The scientists suspected that these oligomers exist only fleetingly—as intermediate aggregates on the way from single-copy “monomers” to the long “polymer” fibrils. The oligomers also were very difficult to study in experiments, because they were biochemically indistinguishable from monomers and fibrils. Thus they were largely ignored.

Then in 1998, William Klein at Northwestern University showed that A-beta, when forced to stay in oligomer form, immediately impairs neuronal synapses and eventually kills neurons, even if A-beta fibrils don’t. Throughout the 2000s, evidence accumulated that oligomers are the only truly toxic forms of A-beta, not just in the lab dish but in animal models. Researchers also found evidence that other amyloid-forming proteins, such as tau, Parkinson’s disease-related alpha synuclein protein, Huntington’s disease-related huntingtin protein, and even the PrP protein implicated in prion diseases, are toxic to neurons principally as oligomers, not as large amyloid fibrils.

These oligomers seem to share a common structure or “conformation” that makes them toxic somehow. A number of labs including Charles Glabe’s, at the University of California–Irvine, have made conformation-specific antibodies that can recognize toxic oligomers of many of these proteins despite their very different amino-acid sequences. Glabe and others also have found evidence that the most toxic oligomers may be those that exist in stable ring-like shapes, and don’t go on to become fibrils.

The idea that oligomers are the toxic aggregates in both Alzheimer’s and prion diseases has helped to reunite the two fields of research. There has also been an acceptance of the likelihood that Alzheimer’s does have a genuine, if weak, prion-like transmissibility. Matthias Jucker at the University of Tubingen showed in 2006 and 2010 that Alzheimer’s pathology, in effect, can be transmitted to mice by injecting Alzheimer’s brain extracts into their brains or bodies.

No one knows why A-beta aggregates seem so much less transmissible than the PrP aggregates in prion diseases. Some scientists suggest, for example, that PrP aggregation can be accelerated by interactions at cell membranes that don’t occur for A-beta. “It’s speculative at this point,” Wisniewski says.

In any case, over the past several years, other experimenters have found evidence that aggregates of tau, huntingtin, alpha-synuclein, and the Lou Gehrig’s disease-associated protein superoxide dismutase 1, all can spread—in prion-like fashion—from cell to cell, in the lab dish or in mouse brains. In recent clinical trials of neuronal transplants, alpha-synuclein aggregates appeared to spread from Parkinson’s patients’ brains into healthy, transplanted neurons.

There is evidence, too, that aggregates of one of these proteins can act as seeds or templates that initiate the aggregation of others. Some researchers suggest that this cross-seeding may sometimes be a trigger or accelerant of neurodegenerative diseases. Certainly for Alzheimer’s disease, there has long been circumstantial evidence that the buildup of A-beta aggregation over decades eventually triggers tau aggregation—which represents the last, lethal stage of disease. There is some experimental evidence, too, that A-beta oligomers can directly induce tau oligomer formation in the lab dish. Dennis Selkoe at Harvard Medical School (also Dana Alliance) also showed last year that A-beta oligomers harvested from human Alzheimer’s brains can trigger the formation and spread of tau tangles in cultured neurons.

How these oligomers harm synapses and kill neurons, and why certain oligomeric proteins kill some neuronal types and not others, are among the major unanswered questions. Whether the initial oligomers always arise spontaneously within the brain, or sometimes come infectiously from outside (see “Does Parkinson’s Disease Start Outside the Brain?”) is another unresolved issue.

Researchers also want to know more about the dynamics of fibril and oligomer self-propagation within tissues. (See story in top right column, "How Alzheimer's may spread" ) But even in advance of knowing the detailed answers to these questions, researchers have established basic treatment strategies: Most aim to boost systems in the brain that naturally reduce toxic aggregates, or to deliver antibodies or other agents that directly remove toxic aggregates. “Ideally, you want to target the abnormal oligomeric structures,” Wisniewski says.