Through The New York Academy of Sciences’ popular discussion groups, doctors and scientists are able to advance our understanding of diseases like Alzheimer’s and Parkinson’s.
Published September 1, 2006
By Alan Dove
Academy Contributor
It’s a favorite of chemistry teachers around the world: the seed crystal demonstration. The instructor drops a grain of salt into a beaker holding a supersaturated solution. Patterning its growth on this initial seed, the rest of the salt in the solution begins to crystallize, extending delicate spikes throughout the container. With any luck, the demonstration itself serves as another type of seed, crystallizing an important scientific concept in the students’ brains.
Crystallization is an intuitive metaphor for learning, but it might not be entirely metaphorical. In recent years, neuroscientists have discovered that a strikingly similar process, the formation of aggregates of specific proteins, could be critical for both normal brain biology and a whole spectrum of neurodegenerative diseases.
“A very common theme in neurodegeneration is that certain proteins aggregate where they shouldn’t, usually in neurons and axons and so forth, and this leads to neurodegeneration,” says Michael Wolfe, associate professor in the Center for Neurologic Diseases at Harvard Medical School.
Exactly how those proteins aggregate, and how new therapies could stop that process, are among the major topics for the Academy’s Neurodegenerative Diseases discussion group, one of the most popular discussion groups in the Frontiers of Science series. Not all neurodegenerative diseases involve protein clustering, but the phenomenon is strikingly common in the field, as recent discussions at events for The New York Academy of Sciences (the Academy) have shown.
A Tangled Web
In neurodegenerative disease research, perhaps no name looms larger than that of Alois Alzheimer, whose 1906 description of a distinctive form of dementia set off a century of discoveries and debates. One of Alzheimer’s central discoveries, the amyloid protein plaques that appear between neurons in the brains of some demented patients, has also been one of the biggest bones of contention in the field. Are these amyloid plaques causing Alzheimer’s disease, or are they merely side effects of other pathological processes?
While that distinction is tremendously important for understanding the basic mechanisms of Alzheimer’s disease, researchers looking to treat the disease have a more pragmatic view. “There are many different arguments about whether amyloid plaques cause Alzheimer’s disease…but I think all agree that [plaque formation] is associated with Alzheimer’s disease, and having an ability to identify that as early as possible is certainly going to open up new doors as far as diagnosis and treatment are concerned,” says Washington University’s Mark Mintun, who spoke at the Academy’s “Imaging and the Aging Brain” conference in May.
Indeed, even the basic researchers consider the amyloid debate passé. “With respect to Alzheimer’s disease there’s a pretty good consensus that amyloid is essential to that process,” says Wolfe, who has helped organize some of the meetings for the Neurodegenerative Diseases group.
Many types of cells throughout the body make amyloid precursor protein, which a protease then cleaves into fragments. The fragment containing the first 42 amino acids, also known as Aβ42, seems to be the main component of amyloid plaques.
Not the Whole Story
But amyloid isn’t the whole story. The other major feature of Alzheimer’s disease pathology is the formation of neurofibrillary tangles within neurons. The tangles are aggregates of the phosphorylated form of another protein, called tau. As Aβ42 aggregates between neurons to form amyloid plaques, phosphor-tau aggregates within neurons to form neurofibrillary tangles. The two processes together disable and destroy the neurons, leading to memory loss.
Unfortunately, researchers face a shortage of good model systems in which to study these processes. Test tube experiments and structural studies have revealed a great deal about the specific interactions that stabilize the protein aggregates, but what happens in living cells, or whole living brains? Humans are the only animals that develop Alzheimer’s disease naturally, so to study the condition in animals, scientists first had to build a better mouse.
One of the best animal models for Alzheimer’s disease research is the cryptically named Tg2576 transgenic mouse, developed by Karen Hsiao Asheand her colleagues at the University of Minnesota medical school. Ashe spoke at the Neurodegenerative Diseases discussion group’s May 23, 2005, meeting and also at a recent symposium cosponsored by the Academy and the Harvard Center for Neurodegeneration and Repair.
Small Aggregates, Good Markers
The Tg2576 mice express a mutant form of the human amyloid precursor protein, and as they age, they develop some of the symptoms of Alzheimer’s disease. Their brains accumulate amyloid plaques, and their performance on learning and memory tests deteriorates, but the animals do not develop neurofibrillary tangles or show a gross loss of neurons.
That profile suggests that the mice might model the earliest stages of Alzheimer’s disease, when patients start to show memory loss long before they lose significant numbers of neurons. Following this lead, Ashe and her colleagues have isolated Aβ from the brains of the mice and characterized it biochemically.
The Aβ molecules seem to cluster in multiples of three, and aggregations of 9 or 12 copies of the protein specifically correlate with memory loss. That suggests that these small aggregates might be good markers for the earliest stages of Alzheimer’s disease, and perhaps good targets for new drugs.
Bench to Bedside to Bench
New drugs and diagnostic tests are hot topics among neurodegenerative disease researchers, and discoveries in this field often move out of the lab and into the clinic very fast. “There’s still a lot of basic research going on, and these are such critical problems that people are just desperate to have treatments,” says Wolfe.
The January 30 meeting of the group featured some impressive examples of this rapid bench-to-bedside translation. At that meeting, Kaj Blennowof the University of Göteborg, Sweden, set the tone in a presentation that ranged from laboratory discoveries about Aβ and tau to a comprehensive nationwide program for the early diagnosis of Alzheimer’s disease.
Blennow and his colleagues focus on three biomarkers in cerebrospinal fluid that correlate with the development of Alzheimer’s disease: the total amount of tau protein, the proportion of tau that is phosphorylated, and the amount of Aβ42. None of these markers is sufficient for a reliable diagnosis on its own, but combining all three can reveal both the progress and the severity of the disease much more accurately than traditional tests.
After fine-tuning the test, Blennow and his colleagues now operate a regular diagnostic service from their laboratory, processing samples of cerebrospinal fluid from nearly half of all patients diagnosed with dementia in Sweden. The testing helps patients and their doctors plan for the disease’s progress, and also provides a critical tool for testing new therapies.
Targeting Tremors
Though Alzheimer’s disease is the most prevalent neurodegenerative condition, Parkinson’s disease, which affects about one million people in the U.S. alone, is another recurring topic for the Neurodegenerative Diseases discussion group.
The disease causes a characteristic pattern of symptoms, including tremor, rigidity, slowed movement, and loss of balance. The problems correlate with a loss of dopamine-secreting neurons in the substantia nigra, a specialized structure in the middle of the brain.
Since the 1960s, doctors have treated Parkinsonism with L-dopa to replace the lost dopamine, but this strategy is unsustainable. Patients need higher and higher doses of the compound over time, until it no longer works. Implanting dopamine-secreting fetal tissue or embryonic stem cells might keep the disease in check, but with current restrictions on research, that treatment is hard to test.
In an effort to find better solutions, researchers are focusing on the molecular basis of the disease, which involves a familiar theme: protein aggregation. In Parkinson’s disease, the surviving neurons in the substantia nigra develop characteristic structures called Lewy bodies, which contain aggregates of the protein α-synuclein. Interestingly, a very closely related protein, β-synuclein, cannot form aggregates.
At the December 8, 2005 meeting of the Neurodegenerative Diseases group, Benoit Giassonof the University of Pennsylvania discussed a clever way of exploiting this difference to learn more about the two proteins. By combining sequences from α- and β-synuclein into chimeric proteins, Giasson and his colleagues defined the specific region responsible for α-synuclein’s aggregation.
“We believe this region in the middle of α-synuclein is the key to making fibrils, and it’s also why β-synuclein cannot make fibrils,” says Giasson. That could give drug developers a well-defined target for the next generation of Parkinson’s disease treatments.
A Parkinsonian Poison
Giasson argues that Lewy bodies and other large α-synuclein aggregates are a major cause of neuronal death in Parkinson’s disease, but in a controversy strikingly parallel to the β-amyloid debate, not all Parkinson’s disease researchers agree. Unlike the β- amyloid controversy, the debate over α-synuclein’s exact role is still unsettled.
For example, Columbia University’s Serge Przedborskiis among those who think that Lewy bodies may be just a marker of the disease, while other mechanisms actually kill the cell. Przedborski presented his view—and his data—at the same meeting where Giasson spoke.
In much of his work, Przedborski relies on a mouse model of a tragic human experiment. In the 1980s, some young, otherwise healthy people began showing up in California hospitals with a bizarre syndrome: neurological symptoms that rapidly progressed to resemble advanced Parkinsonism. These patients had all taken a synthetic form of heroin contaminated with a known industrial toxin, 1-methyl-4-phenyl-1,2,5,6- tetrahydropyridine, or MPTP.
The body metabolizes MPTP to a compound called MPP+, which selectively accumulates in the substantia nigra and kills dopamine-secreting neurons. MPTP has the same effect in mice, providing a valuable animal model for Parkinson’s disease. Przedborski and his colleagues have combined this toxin-based approach with sophisticated mouse genetics to pinpoint the causes of neuronal death in the substantia nigra.
So far, the researchers have uncovered at least four ways MPP+ can kill a neuron. The compound increases the release of dopamine into the cytoplasm, stalls the cell’s energy-producing electron transport system, generates reactive oxygen species, and stimulates inflammation that can cause more damage to neighboring cells. Interestingly, reactive oxygen species may target α-synuclein especially, mimicking an uncommon inherited form of Parkinson’s disease. “You’re altering the properties of these important proteins [chemically], and you imitate what the mutations can do,” says Przedborski.
Don’t Forget the Prions
With all the news about aggregating proteins causing horrific damage to the brain, it would be easy to get the impression that aggregation is always bad. But in fact, it may be an essential feature of normal brain biology. That’s the surprising conclusion attendees heard at the “Imaging and the Aging Brain” conference in May, when Nobel laureate Eric Kandeldiscussed his laboratory’s latest results.
In order to form long-term memories, the brain rewires the synaptic connections between neurons, a process that requires building a new synapse, then ensuring that it persists. In the simple nervous systems of sea slugs, Kandel and his colleagues discovered that even before the construction materials reach a future synapse, the messenger RNA (mRNA) for synapse maintenance is already there, cached in an inactive form. The arrangement is reminiscent of the first stages of egg development in amphibians, in which mRNAs for early embryonic growth are already preformed before fertilization.
The investigators discovered that in the slug neurons, a protein called ApCPEB keeps the cached RNA inactive until the signal arrives to maintain a new synapse. That’s not too surprising, but ApCPEB has an odd feature for a gene regulator: a prion-like sequence at one end.
Dramatic Neurodegenerative Conditions
Prions are infamous for forming the protein aggregates that cause some dramatic neurodegenerative conditions, including bovine spongiform encephalopathy, or “mad cow” disease. Based on his new data, though, Kandel argues that “there is a subclass [of prions] in which the aggregated form is actually the functional form of the protein.”
Though Kandel and his colleagues are still testing the theory of “good” prions, their preliminary results suggest a new, more nuanced view of protein aggregation in the brain, in which protein clustering can kill us, but it may also be essential for survival. If that yin-yang relationship is correct, it will bring the seed crystal metaphor full circle: remembering a grain of salt really might crystallize an idea.
Also read: Transformative Research in Neurodegenerative Disease and Neuropsychiatric Disorders: 2017 Innovators in Science Award Symposium
About the Author
Alan Dove is a science writer and reporter for Nature Medicine, Nature Biotechnology, and Genomics and Proteomics. He also teaches at the NYU School of Journalism.
