How do aggregates of certain proteins, such as amyloid beta and tau in Alzheimer's, cause the deaths of brain cells? Much modern neurodegenerative disease research has been aimed at answering that question.
But one might as well ask how those aggregates fail to kill brain cells in some people.
"There's an old conundrum in neurology," says Harvard researcher Bruce Yankner. "Many elderly individuals die with enough pathology [amyloid beta plaques and tau tangles] for a pathologist to classify their brains as having Alzheimer's-yet there's no evidence that they had dementia."
What's the difference between an elderly brain that has Alzheimer's pathology but normal cognitive function, and one that succumbs to that pathology with massive neuronal losses and dementia?
A protective protein called REST is the difference-or at least a big part of it-according to a study by Yankner and colleagues that was published in Nature in late March. In cognitively normal elderly brains, the protein exists at relatively high levels in neurons, particularly in the brain regions that would be first affected by Alzheimer's. In brains with Alzheimer's dementia, by contrast, REST activity is much lower in those affected regions.
The findings hint that in many people, high levels of REST keep their brains free from Alzheimer's and other neurodegenerative diseases into their 90s and beyond.
"The data we have suggest to us that REST could regulate the tipping point, so to speak, between preserved cognitive function and cognitive decline," says Yankner.
Old neurons need their REST
Yankner and his laboratory were first known for work on the toxicity of amyloid beta-a thorny and controversial issue that hasn't been completely resolved. It now appears that the protein harms neurons principally when it forms very small aggregates (oligomers), rather than the large, tough aggregates that make up the visible amyloid "plaques" in Alzheimer's brains. But precisely how amyloid beta oligomers lead to massive, dementia-causing neuronal losses remains unclear.
About a decade ago, Yankner and his colleagues began, in effect, to cast a wider net for factors relevant to Alzheimer's, by looking at the changes that occur in the aging brain-changes in the patterns of gene expression in neurons, for example.
In their new study, they identified a protein of interest called REST (Repressor Element 1-Silencing Transcription factor). This protein normally works on the DNA of cells to regulate their gene activity. In non-neuronal cells, and in immature neurons, REST suppresses genes that should be turned on only in adult neurons. Yankner's team, including first author Tao Lu, determined that REST also plays a big role in the aging brain. Indeed, of all the genes that work in neurons, REST, they found, is the most likely to become significantly more active as the brain moves from early to late middle age. And among the genes that get switched off in the aging brain, a surprisingly large number appear to be switched off by REST.
Curiously, although REST levels tended to increase with aging in the nuclei of the healthy neurons they sampled, the researchers found that in Alzheimer's brains, the protein was virtually absent from neuronal nuclei in the hippocampal and cortical regions that are most affected by the disease.
Those findings hinted, but did not prove, that REST is a protective factor whose loss leads to Alzheimer's. That case grew stronger when the researchers looked at the genes that REST normally suppresses. These turned out to include "cell death pathway" genes, which facilitate the demise of stressed neurons. Also on the list were genes with direct relevance to Alzheimer's, including genes for enzymes that help make amyloid beta and phosphorylate tau.
This evidence was still only circumstantial. But then the scientists examined what happens to brain cells when REST is absent. They cultured mouse neurons, some normal and some lacking REST, and found that the latter were much easier to kill with stressors such as hydrogen peroxide (the classic oxidative stress molecule) and amyloid beta oligomers. The weakness of these REST-deficient neurons could be reversed by using a gene-therapy type approach to boost REST levels.
The researchers observed similar consequences in REST-deficient transgenic mice: At one month of age their brains seemed like those of normal mice; but by eight months (a third of the way through a typical lab mouse lifespan) showed unmistakable marks of neurodegeneration, with substantial losses of neurons in the cortex and hippocampus, and activation of inflammatory microglial cells.
To round out the animal studies, the researchers showed that a distant evolutionary cousin of mammalian REST, a transcription factor called spr-4, has a similar role in resisting oxidative and amyloid-beta stress, and preventing early death, in C. elegans worms.
Finally, to underscore the relevance to humans, Yankner's team examined samples of human brain tissue from several large-scale Alzheimer's-related studies. They found that higher REST levels in the nuclei of prefrontal and certain hippocampal neurons correlated strongly with better cognition-and with cognitive preservation in brains whose plaque-and-tangle pathology otherwise met criteria for Alzheimer's.
Higher REST levels in prefrontal neurons also correlated with longevity. Indeed, when Yankner's team looked at the brains of people who had died at ages of 100 and above, they found unusually high levels of the protein in prefrontal and hippocampal neurons-again suggesting strongly that REST has a powerful ability to protect the brain from the degenerative forces of aging.
Why does REST disappear in Alzheimer's?
The findings have been noted throughout the field of neurodegenerative research. "Dr. Yankner used many different ways to validate his conclusion that REST levels correlate with the health of the brain," says Li-Huei Tsai, a researcher at the Massachussetts Institute of Technology and member of the Dana Alliance on Brain Initiatives (DABI) who co-authored a commentary in Nature on the Yankner paper.
But some big questions remain. One, of course, is whether boosting REST levels in the nuclei of neurons in middle aged people will really ward off Alzheimer's and help extend lifepsan.
Another big question is how to accomplish such a boosting-because it isn't yet clear why REST is depleted from neuronal nuclei in Alzheimer's-affected brain regions.
One clue has to do with a growth and survival related cell signaling pathway called the Wnt pathway. Yankner's team found that boosting Wnt signaling boosts REST levels, whereas reduced Wnt signaling-normally seen in affected regions in Alzheimer's-correlates with reduced REST levels.
"Wnt signaling is a potential mechanism in regulating REST levels in the brain," says Tsai. "But then it's a chicken and egg question, because what regulates Wnt signaling?" Even the sources of Wnt among brain cells, she notes, have yet to be completely understood. In any case, rises in Wnt signaling have been linked to various cancers including brain cancer, so simply boosting Wnt in an elderly brain could do more harm than good.
Another clue to REST depletion in Alzheimer's has to do with a process called autophagy-an internal garbage-disposal process on which aging neurons are thought to rely heavily. The buildup of unwanted protein aggregates, such as amyloid beta and tau oligomers, appears to trigger an increase in autophagy in neurons. But Yankner's team found that in neurons from Alzheimer's affected regions, waste-digestion sacs called autophagosomes had accumulated and were filled not just with Alzheimer's-related protein aggregates, but with REST proteins too.
One hypothesis, which Yankner and his colleagues plan to investigate, is that Alzheimer's aggregates spur increases in autophagy that can, at high enough levels of activity, keep new-made REST proteins from moving normally into the nucleus.
On the other hand, Ralph Nixon, a researcher at New York University's Langone Medical Center and the Nathan Kline Institute, and also a DABI member, finds it "anti-intuitive" that autophagy, meant to be a protective response, would end up shutting down another powerful protective mechanism such as REST. Nixon's work suggests that autophagy isn't just upregulated in Alzheimer's-it also becomes dysfunctional, so that waste isn't digested properly, and builds up, thus potentially explaining the abnormal accumulation of autophagosomes. "I would suggest that the problem could be the failure of autophagy and the subsequent failure of signals, emanating from digestion, that should regulate the translocation of REST to the nucleus," he says.
To resolve such questions, and to get better insights into therapeutic strategies, Yankner and his team are now looking at what happens to Alzheimer's mouse models when their neurons have different levels of REST. "We're going to see whether varying levels of REST in the setting of neurodegenerative disease pathology can overcome what we perceive as this mislocalizing force [on REST]," he says.
The work, Yankner notes, could have implications beyond Alzheimer's. In the study, his team looked at the brains of people who had died with fronto-temporal dementia, a tau-related disease, and Lewy-body dementia, an alpha-synuclein-linked disease. In each series, just as in the previously studied Alzheimer's brains, REST levels on average were much lower than normal in affected cortical neurons. The researchers also found abnormal accumulations of REST outside neuronal nuclei in autophagosomes, again in association with disease-linked protein aggregates. Thus, although much work still has to be done, there is the possibility that a REST-boosting drug could protect aging brains generally against neurodegenerative disease.