Possible Alzheimer's gene stirs conflict.

It's naive to think that scientists aren't as combative as athletes or businesspeople. And nothing gets the competitive juices flowing as quickly as the hunt for a gene responsible for a feared human disease.

Since last year, investigators have been racing to find a gene on chromosome 12 that they suspect influences a person's chance of getting Alzheimer's disease late in life. At a meeting in Amsterdam this week, and in the August NATURE GENETICS, one group of researchers has fingered as the culprit an altered form of a gene called A2M.

"If you have the A2M mutation, this puts you at risk for Alzheimer's," says Rudolph E. Tanzi of the Massachusetts General Hospital in Boston, who led the study.

At the same meeting, however, two rival research teams rejected that claim. "We can't support it," says Margaret A. Pericak-Vance of Duke University Medical Center in Durham, N.C.

"Our own data don't show the same association, so the most parsimonious explanation is that it's not A2M but something nearby," agrees Peter H. St. George-Hyslop of the University of Toronto.

Tanzi counters that the two groups haven't adequately replicated his study. Moreover, he believes that some of the other scientists' results actually lend support to his group's conclusions. "I'm really surprised at how they're interpreting their data," Tanzi told SCIENCE NEWS. "What you have here is the strategic use of positive and negative data, taken out of context, to knock this result."

Everyone agrees that A2M is an appealing candidate for a gene involved in Alzheimer's disease. It encodes alpha-2 macroglobulin, a protein that deactivates proteases, enzymes that carve up other proteins.

Alpha-2 macroglobulin attaches to the same cell surface protein as APP and APOE, the products of two other genes involved in Alzheimer's disease. Moreover, studies have indicated that the unmutated form of the protein helps get rid of beta-amyloid, the APP fragment that accumulates excessively in the brains of Alzheimer's patients.

Initial genetic analyses, in which scientists compared the DNA of elderly people with Alzheimer's disease to unrelated, age-matched, healthy individuals, didn't confirm A2M's link to the disease, admits Tanzi. His group then conducted sib-pair analyses, which compare siblings with and without a disease. The strategy, says Tanzi, reduces confounding factors of traditional linkage studies, such as the varying genetic background of those examined.

Using such analyses, Tanzi and his colleagues found that about 17 percent of people who got Alzheimer's after age 85 had a particular mutation in A2M. That same mutation appeared in only 4 percent of the siblings who were 85 or older and didn't have the disease.

Tanzi suggests that the mutant form of A2M makes the aging brain more vulnerable to Alzheimer's disease, although it doesn't guarantee that the illness will strike. It's unclear how the identified mutation, a small deletion, affects alpha2 macroglobulin. It may hamper the protein's ability to bind other proteins.

As they tussle over A2M, researchers are also reassessing the gene that encodes APOE. This gene comes in three versions: E2, E3, and E4. Having the E4 version was thought to increase a person's odds of getting late-onset Alzheimer's, but a new study, reported in Amsterdam and also in NATURE GENETICS, suggests that the various APOE forms influence when, rather than if, a person gets the disease.

John C.S. Breitner of Johns Hopkins University in Baltimore and his colleagues found that about 40 percent of people, regardless of which APOE variant they had, got Alzheimer's disease if they lived long enough. This observation, notes Breitner, was likely obscured in earlier studies because they failed to include enough very old people, he says.