Researchers who surveyed the DNA of thousands of Icelanders have found a rare gene variant whose carriers live longer, almost never get Alzheimer’s disease, and show markedly better cognitive function even at advanced ages. The finding appears to validate the idea that reducing the production of the Alzheimer’s-linked protein amyloid beta (Aβ) could prevent Alzheimer’s. It also suggests that Aβ is a large but hidden cause of the cognitive decline observed in elderly people who haven’t been diagnosed with Alzheimer’s.
Future drugs that can recreate this Aβ-reducing effect “should perhaps be given not only to people at risk of Alzheimer’s but to all elderly people,” says Kári Stefánsson, senior investigator of the study, which appears online in Nature today.
Stefánsson is the CEO of deCODE Genetics, a genomics and biotechnology company that he founded in Reykjavik, Iceland, in 1996, after a career in academic neurology and neuroscience at Harvard Medical School and the University of Chicago.
In the new study, Stefánsson’s team surveyed the genomes of 1,795 elderly Icelanders. They looked specifically for unusual forms of the gene that codes for amyloid precursor protein (APP), the neuron-produced “mother protein” from which Aβ is cleaved by enzymes. Scientists already know that some rare mutants of APP result in the production of more Aβ (or more aggregation-prone Aβ), and cause an inherited, early-onset form of Alzheimer’s. In this case Stefánsson wanted to find mutations that affect the risk of the more common, late-onset form of the disease.
Surprisingly, his team found a rare APP gene mutation, designated A673T, that seems to strongly protect people from Alzheimer’s—the first ever that has been shown to do so. In their sample, those who had lived to an age of at least 85 years and remained “cognitively intact” according to standard tests were about seven and a half times more likely to have the mutation than the over-85s who had Alzheimer’s. The researchers confirmed these results in larger genome databases, from which it appeared that APP-A673T protects almost completely against Alzheimer’s. “Out of several thousand Alzheimer’s patients, we found five who carried the mutation, and one of them died at the age of 101; another died at the age of 100; a third at the age of 95,” Stefánsson says.
Even among elderly Icelanders who did not have an Alzheimer’s diagnosis, those who carried the A673T variant seemed to enjoy a striking cognitive advantage. A sample of nursing-home carriers at age 90 scored much better on cognitive tests than the average 80-year-old non-carrier. Sample sizes for carriers were small—the mutation occurs in only 0.5 percent to 1 percent of Scandinavians—but the data hint that the A673T mutation delays by one to two decades the cognitive decline normally experienced by the elderly. “This is extraordinarily important, because it indicates that Alzheimer’s disease and ‘normal’ cognitive decline are on the basis of the same mechanism,” says Stefánsson.
It’s all about amyloid beta
What is that mechanism? Stefánsson’s team inserted the gene for APP-A673T into test cells and found that the mutant APP protein resists the cleavage by enzymes that normally releases Aβ. The result was a near-halving of normal Aβ production. (Researchers widely believe that higher Aβ production raises the risk that small, toxic aggregates of the protein will form.) The effect, they found, was specifically based on mutant APP’s relative inability to be cleaved by the enzyme beta secretase. Also called BACE1, this enzyme makes the first of the two cuts on APP that release Aβ.
Higher than normal levels of BACE1 are considered a risk factor for common, late-onset Alzheimer’s. Anything that enhances BACE1’s ability to cut APP also is considered a likely risk factor. Another previously described APP mutation, A673V—an alteration at the same point on the gene where A673T occurs—has this BACE1-enhancing effect, and results in higher levels of Aβ, as well as early-onset Alzheimer’s in people who have two copies of the mutant gene. As Stefánsson’s group notes in their paper, changes at this highly sensitive amino-acid position #673 appear to be able to increase or decrease BACE1 cleavage of APP, with profound effects on Alzheimer’s risk, old-age cognitive ability—and even longevity, since Alzheimer’s is both common and fatal.
Researchers have known about BACE1’s role in the production of Aβ since the 1990s, and pharmaceutical companies have been developing drugs that inhibit the enzyme. A problem with such inhibitors is that they can inhibit other enzymes besides BACE1, causing adverse side-effects; in addition, they don’t easily cross from the bloodstream into the brain. Some researchers are developing an anti-BACE1 antibody therapy, but it, too, does not cross into the brain easily. Even a therapy that precisely targets BACE1 may have side effects: There is evidence that the enzyme works on other proteins in the brain besides APP, and interfering with those processes could be harmful.
The new finding at least provides support to the much-debated principle that inhibiting Aβ production should be a major goal of Alzheimer’s drug development. “This gives the amyloid hypothesis a nice shot in the arm,” says Sam Gandy, a professor of neurology and psychiatry who directs the Center for Cognitive Health at the Mount Sinai School of Medicine.
The major caveat of the Aβ reduction strategy is that it should be done early in the disease process, before other, less-reversible degenerative processes take hold. But in this early phase people don't have obvious cognitive symptoms: The only signs that they are developing disease may be abnormal brain-scans for insoluble deposits of Aβ, and abnormal levels of Aβ and the related protein tau in cerebrospinal fluid samples. The absence of clinical illness in people who have these signs creates ethical and cost-related complications for clinical trials, which effectively must be “dementia prevention” trials. (See “Alzheimer’s prevention trials set to start in 2012.”) “Considering how long and difficult prevention trials are likely to be, this [new study] could not have come at a better time,” says Gandy.
Does Aβ explain normal age-related cognitive decline?
One of the most surprising findings in the new study is the linkage of the A673T mutation to better cognitive scores among elderly people who don’t have Alzheimer’s. This hints that A673T protects not only against Alzheimer’s but also against a large part of what neurologists consider normal age-related cognitive decline.
This normal, slow, “non-Alzheimer’s” cognitive decline is often seen as a gradual decrease in efficiency in a wide variety of brain areas—for example, the prefrontal cortex areas that mediate working memory. Stefánsson suggests that the two conditions are more closely related than had been thought—in other words, that Aβ could be a major cause of cognitive decline among elderly people, even when they don’t have a diagnosis of Alzheimer’s.
Alzheimer’s research has been pointing in this direction, especially in recent years, as new brain-imaging technology has allowed scientists to detect Aβ deposits long before cognitive symptoms arise. “About thirty percent of people who are age 70 and cognitively normal already have the [Aβ-based] brain lesions of Alzheimer’s,” says John C. Morris, a professor of neurology, pathology and immunology who directs the Alzheimer’s Disease Research Center at Washington University at St. Louis. The implication is that many elderly people who do not have Alzheimer’s are nevertheless in a preclinical stage of the disease process, in which they accumulate Aβ and move—perhaps very slowly—towards dementia, which would be diagnosed if they were to live long enough. “My hypothesis here,” says Morris, “would be that if in fact this mutation in the APP gene protects against Alzheimer’s, it may also be operating in the same beneficial way for these people with preclinical Alzheimer’s.”
Stefánsson plans more detailed studies of Aβ levels in people who carry the mutation and the cognitive differences between them and similarly-aged non-carriers. His company is using genomics techniques to help develop treatment strategies and diagnostic tests for dozens of human diseases, but thanks to Stefánsson’s background in neurology, deCODE has a strong emphasis on brain ailments. The company’s projects cover Parkinson’s disease, schizophrenia, epilepsy, pain, autism, dyslexia, and Tourette’s syndrome. In recent years, they also have been sifting through human genomes to relate gene variants to normal variations in cognitive abilities—“as a first step in using genetics to figure out how the brain thinks,” Stefánsson says.