By David N. Leff

It's common knowledge by now that senile neuritic plaques are the defining hallmark of Alzheimer's disease (AD).

These aggregations of amyloid protein clump around the neurons that dwindle with advancing AD, and can only be seen by brain biopsy or postmortem autopsy.

AD plaques consist of a peptide called amyloidb (Ab), about 39 to 42 amino acids long and derived from a much longer Ab precursor protein. In vitro, at least, the peptide is extremely insoluble, and precipitates to form the characteristic plaque aggregates.

What makes them behave this way, and how culpable the plaques are of causing AD, are among the many puzzles that enshroud the mystery of the disease. (See BioWorld Today, Feb. 6, 1997, p. 1.)

"Nobody knows what Alzheimer's is," said polymer physicist H. Eugene Stanley. "We only know its symptoms, and this is the kind of problem that attracts us physicists, who like to feel that we can figure out puzzles."

One clue that attracted him and his lab crew of statistical physicists at Boston University to the amyloid plaque puzzle was the peptide's proclivity for aggregating. "In our branch of polymer physics," Stanley told BioWorld Today, "we specialize in aggregation and disaggregation of various objects, polymers, monomers or more complicated forms of matter."

He is corresponding author of a paper in the current Proceedings of the National Academy of Sciences (PNAS), dated July 8, 1997. Its title: "Aggregation and disaggregation of senile plaques in Alzheimer's disease." The article's senior author is clinical and research neurologist Brad Hyman, who directs the Alzheimer's Research Unit at Boston's Massachusetts General Hospital.

Using the latest-model tools of their esoteric trade, Stanley and Hyman report two discoveries that make amyloid plaques seem less immutable than commonly thought: They are porous rather than solid, and their pattern of aggregation suggests that potentially they may disaggregate as well. "In other words," Stanley said, "the plaques are full of holes. Also, the plaque is not something that just assembles itself in the brain and remains frozen for all time. Rather, it assembles itself and can then disassemble, or come apart."

But he hastened to add: "Don't get too excited. We didn't say that senile plaques can be dissolved, and therefore Alzheimer's can be cured. What our experimental data do show," he added, "is that there is a dynamic steady-state equilibrium in plaque deposition, which means things going back and forth in both directions -- aggregation and disaggregation."

What makes this finding exciting, Stanley explained, "is that if we could figure out the parameters that control this equilibrium, one might shift it in the direction of less aggregation, more disaggregation, and then we'd obviously have a potential cure. That's encouraging," he pointed out, "because otherwise it could mean that plaques assemble and stick for good."

If Arterial Plaques Can Reverse, How About Senile Ones?

Stanley defended this mitigated optimism by comparing AD plaques with the atherosclerotic plaques (no biochemical relation) that cause coronary heart disease and stroke. "At one time," he recalled, "it was thought that arterial plaque formation was unidirectional. That is, plaque aggregated in the arteries, and that was the end of the story. Now we know that there are drugs, diet, exercise and other things that can encourage those plaques to dissolve themselves. So we began to develop a picture of a senile plaque as a more dynamic thing."

That picture developed on the screens of a 3-D laser confocal microscope and a versatile imaging computer monitor.

The first exercise was to confirm the group's hunch that plaques are porous. Two years ago, they described a graphic histogram that sorted plaques by size. "It showed a most unusual geometric curve," Stanley recounted, "one it would never have had if the plaques were solid objects. The graphic ended with a long tail -- like a bell-shaped curve in a high wind."

Stanley added: "It's a novel contribution, as far as I know -- the best evidence that real plaques are porous objects, with pore sizes in the neighborhood of a few microns."

The co-authors examined brain tissue from six deceased AD patients under their confocal microscope. This instrument produces 3-D images, like a CAT (computer-assisted tomography) scanner, but uses laser light instead of X-ray. The team "sliced" the specimen plaques at one-micron intervals, to build up 3-D reconstructions. Roughly spherical in shape, typical plaques were 60 microns in diameter.

"Because light has the ability to resolve such tiny dimensions, about the wavelength of light," Stanley observed, "we could see very tiny structures in three dimensions. That's what's cool about this technique."

Then the group constructed a computer model of the microscope-delineated cerebral plaques. "We wanted to visualize it on the screen," recalled statistical physicist Luis Cruz, the PNAS paper's first author. "So we created in the memory of the computer a volume -- which is a 3-D array -- where we could paint the digitized pixels [2-D 'picture elements'] inside."

Cruz continued: "We initially started with a very small amount of occupied pixels in the center of the image, then applied certain rules of aggregation and disaggregation at the molecular level, which the computer can repeat thousands of times.

"We translated the actual sticking of the protein as an aggregation, disaggregation as any biological process. At the end of the day, we could calculate from 1,000 to 10,000 steps in a system that evolves to some kind of equilibrium."

Hyman told BioWorld Today: "The implications of the model -- which is only a model -- or at least one possible explanation of the geometric properties that we found, is that the aggregation and disaggregation phenomenon is going on. And that fits with a variety of other indirect evidence that plaques may not be permanent structure.

"Really," he observed, "we're at a very early point in understanding the physical properties of A-beta, and how in vivo A-beta molecules come together to form a big macromolecular structure. And I think one of the messages here is that technology now exists to try to study that process."

On that score, the Boston University physicists have just acquired "the very best imaging computer that exists," Stanley said, "so we will now be able to transfer plaque material from the real brain to the monitor, and not only rotate it around, but look inside of its pores by virtual reality."

He added: "We want to see the correlation between the senile plaques and the absence of neurons. This would be a proof -- now an open question -- that the plaques destroy the neurons."

"If in fact the body has the capability of removing plaques," Hyman concluded, "either passively or actively, then understanding the mechanisms involved in that provide some targets to manipulate the pathophysiology of the disease. *