A surprising role for Alzheimer proteins?

One fact about Down's syndrome, which stems from inheriting an extra copy of chromosome 21, has intrigued Huntington Potter for years. If the person lives long enough, "every Down's syndrome individual gets Alzheimer's disease," notes Potter, a geneticist at the Harvard Medical School in Boston.

That morsel of information, along with several other clues, has prompted Potter to suggest that some, perhaps many cases of Alzheimer's disease result from an abnormal accumulation of chromosome 21 in cells.

While probing cells for the homes of two proteins implicated in inherited forms of Alzheimer's disease, Potter appears to have found more support for his heretical proposal. The proteins reside in centrosomes, kinetochores, and nuclear membranes--structures involved in the dispersal of chromosomes in dividing cells--he and his coworkers report in the Sept. 5 Cell. "It's the first clear data in favor of the chromosome mis-segregation model," says Potter.

"It's a really interesting hypothesis. It's going to be somewhat controversial," predicts Bruce T. Lamb of Case Western Reserve University in Cleveland.

Potter's group studied two proteins called presenilin-1 and presenilin-2. If a person inherits two mutant copies of either presenilin-encoding gene, the abnormal proteins cause Alzheimer's disease to strike early, often before age 50 (SN: 3/2/96, p. 134).

Since that discovery, investigators have struggled to learn the normal cell functions of presenilins. For example, their structure suggests that they are embedded in one or more of the cell's various membranes, but which ones?

Several studies with antibodies that bind to the presenilins have indicated that they reside in the endoplasmic reticulum (ER), where the cell synthesizes proteins, and in the Golgi complex, where proteins are often modified.

Yet Potter contends that those studies misled investigators because they typically involved cells forced to overproduce presenilins. In such cells, presenilins "tend to pile up in the ER and Golgi," he says.

When Potter and his colleagues, working with antibodies they created, studied cells that produce normal amounts of presenilins, they found that the proteins were part of the nuclear membrane, the sac that surrounds a cell's DNA. The scientists also detected presenilins in kinetochores, which are specialized protein complexes on chromosomes, and in centrosomes, structures found just outside the nuclear membrane.

Kinetochores and centrosomes play crucial roles in distributing identical sets of chromosomes within a dividing cell, says Potter. The two centrosomes organize filaments, called microtubules, along which the chromosomes travel to opposite sides of the cell. The kinetochores are the sites at which the chromosomes attach to the microtubules.

In a nondividing cell, suggests Potter, presenilins stud the inner surface of the nuclear membrane and hold onto chromosomes via kinetochores. If a cell begins to divide, presenilins may aid the process by releasing the chromosomes.

While Potter believes that his data provide strong evidence that presenilins participate in chromosome segregation, he notes that it remains unclear how their mutant versions cause Alzheimer's.

Some brain cells in Alzheimer's patients may accumulate abnormal numbers of chromosomes, including chromosome 21, ultimately causing them to die or otherwise falter, he says.

Alternatively, the presenilin mutations may induce cell death by stimulating neurons, brains cells that normally don't divide in adults, to try to divide, notes Peter Davies of Albert Einstein College of Medicine in New York. Davies published data last year hinting that brain cells in Alzheimer's patients may be attempting to split inappropriately.

Although Potter's data and interpretations are likely to be challenged, Davies welcomes the new work. "I love it. It generates a lot of testable ideas and new hypotheses that take us into ground we haven't covered," he says.