Variety is called the spice of life, but it is more like the main ingredient when it comes to the underlying biology. Proteins, for example, often appear in different forms that seem to have similar function but reveal key differences when circumstances change. By selecting from this store of variations, or polymorphisms, evolution arrives at many of its adaptations.
Polymorphisms are central to understanding the origins not just of beneficial traits but of many diseases, most recently late-onset Alzheimer's, according to a new study by HMS and HSPH researchers. Rudolph Tanzi, Deborah Blacker, and their colleagues have found that people carrying a common variant of a protein, called alpha2 macroglobulin (A2M), are three times more likely to develop Alzheimer's disease than those with the standard form of the protein.
A More Potent Risk Factor
Until recently, the only established genetic risk factor for late-onset Alzheimer's was a variant of apolipoproteinE called ApoE4. Risk was seen to increase in people carrying two copies of the gene for this variant. Yet a single copy of the gene for the newly discovered A2M variant increases risk to the same degree, report the researchers in the August 1 Nature Genetics, suggesting that this gene variant, called A2M-2, may be an even more powerful late-onset Alzheimer's risk factor than ApoE4.
| Marilyn Albert, Deborah Blacker, Linda Rodes, and Rudy Tanzi collaborated with others to find the A2M-2 gene. |
Tanzi speculates that while the A2M-2 variant may be able to carry out its main task of degrading amyloid-beta protein (A-beta), the main constituent of Alzheimer's plaques in the brain, it may perform less efficiently than the normal version. "So A2M-2 might allow more A-beta to collect at the synapse," says Tanzi, an associate professor of neurology at Massachusetts General Hospital. As a person ages, the disadvantage could show up in the form of more accumulated A-beta in the brain.
Still, people carrying the gene for A2M-2 do not necessarily develop Alzheimer's. "You can be 90 and carry the A2M-2 allele and not get the disease," says Tanzi. He believes that A2M-2 may work in combination with other genetic variations to bring about the disease. "I think we're going to see that late-onset Alzheimer's involves a combination of effects, all modest, of a large set of genes. So we have to start rethinking the whole paradigm of late-onset Alzheimer's," he says.
The new paradigm--a mixture of genetic polymorphisms that are only revealed to be harmful under conditions of advanced age--could explain why so many older people suffer from Alzheimer's. More than 30 percent of Americans over the age of 85 have the disease. As people live longer, additional harmful polymorphisms could be exposed. "As life spans are extended more and more, how many so-called benign common polymorphisms are just lying in wait and are going to spring up to challenge whether you're going to be able to live past a hundred without getting some form of cancer, diabetes, or stroke?" says Tanzi.
Recognizing Ethnic Variation
When Tanzi and his colleagues began their quest for a new late-onset Alzheimer's gene, the field was littered with unsuccessful candidates. Many were doomed to fail because they were isolated by a faulty technique. Typically, researchers would compare the genes of Alzheimer's patients to those of age-matched unaffected people without paying attention to the ethnic composition of the two samples. Yet ethnic groups vary notoriously in the frequency of certain genetic variants. "What researchers didn't realize is if you see an allele frequency difference, it was most likely because the ethnicity of the cases and controls was not the same," says Tanzi.
| Normally, the A2M protein protects against Alzheimer's by delivering A-beta into the neuron where it is degraded. However, to enter the neuron by the LRP receptor, the A2MA-beta complex must first bind a protease. Tanzi speculates that while the A2M-2 gene is capable of making normal protein, it might produce some fraction that is unable to bind protease. With entry into the neuron impeded, A-beta could accumulate in the extracellular spaces, eventually forming destructive Alzheimer's plaques. ApoE may hasten the disease process by competing with A2M-2 for LRP and A-beta. |
Using family data from the National Institute of Mental Health, he and his colleagues compared Alzheimer's patients with unaffected siblings. Scanning the entire genome for possible areas of difference, they hit upon a region on chromosome 12 that held the gene for A2M. The A2M protein was known to share the LRP receptor with ApoE and also with APP, a protein associated with early-onset Alzheimer's. A2M was also known to play an important role in the brain, regulating the activity of several proteins, including proteases, which degrade damaged or unneeded proteins.
Intrigued, the researchers began looking for polymorphic versions of the A2M gene in the general population. "We found this weird deletion in the database," says Tanzi, referring to the A2M-2 gene, which has five fewer bases than the standard version. To the researchers' surprise, 30 percent of the population carried the A2M-2 variant.
Meanwhile, researchers were uncovering more about the protein. Professor of neurology Dennis Selkoe (at HMS and BWH) and his colleagues found that A2M, when bound by a serine protease, was able to degrade A-beta in vitro. Other researchers were reporting that the A-betaprotease complex could bind LRP. "So all these biological scenarios just all fell out. It was an amazing convergence," says Tanzi.
Focusing on Families
The linchpin turned out to be a powerful new statistical technique developed by Nan Laird, the Henry Pickering Walcott professor of biostatistics at HSPH, with help from Blacker, an assistant professor of psychiatry at HMS and MGH and an assistant professor in the Department of Epidemiology at HSPH. Previously, to compare two siblings required also having data from parents--a rarity among late-onset Alzheimer's patients. Laird and Blacker invented a version of the method that eliminated this need--and, in fact, required only two family members, an affected and unaffected sibling. "That's what changed the world for us," says Tanzi.
He and his colleagues ran the family-based program, and other tests including an odds ratio. "By association and odds ratio, A2M-2 was at least as strong a risk factor as ApoE4 for Alzheimer's disease," says Tanzi. He and others believe that A2M-2 and ApoE4 may interact, the first deciding whether a person will develop Alzheimer's disease and the second when it occurs (see figure).
Much remains to be learned about the latest Alzheimer's gene before it can be used to predict risk for the disease. "We don't know even 5 percent of what we need to know before we begin thinking of using it for that purpose," says Tanzi. "But we do know something new and valid and very useful, which is a healthy functioning A2M is probably protective against age-related changes leading to Alzheimer's."
--Misia Landau
Focus, August 14, 1998