By David N. Leff

Peer deep into the brain cells of a person who died with late-stage Alzheimer's disease (AD) and what do you see? Hordes of senile neuritic plaques wrapped around the afflicted neurons attacked by the disease.

These amyloid plaques constitute the post-mortem diagnostic hallmark of last resort, which sets AD apart from other dementias of advanced age. (See BioWorld Today, Nov. 23, 1999, p. 1.)

Inside those same embattled neurons lurks another AD hallmark - neurofibrillary tangles. These snarled skeins of tau protein are less notorious than plaques, only because they are less understood.

"There are several ways to detect the tangles inside the neurons," observed molecular neuroscientist Li-Huei Tsai, at Harvard Medical School and the Howard Hughes Medical Institute in Boston. "Sometimes the neuron is chock-full of tangles," she went on. "They form a filamentous, aggregated kind of structure, which can occupy the entire cytoplasm of the cell body."

Neurofibrillary tangles consist mainly of a protein named tau (for the 19th letter of the Greek alphabet). "We believe that tau's normal function is to stabilize microtubules in neurons," Tsai explained. "Microtubules are major cytoskeleton proteins that maintain a cell's shape and integrity. Normal tau is more or less localized to the message-transmitting axonal projections of neurons. But abnormal tau accumulates in the cell bodies of late-stage AD brains, and you don't see it in the axons any more."

What makes tau abnormal is what Tsai termed a "phosphorylation binge" - non stop overdosing on phosphate groups plugged into other proteins. She defined phosphorylation as "a very fundamental biological process, which can do a lot of different things. Sometimes phosphorylation turns on other enzymes, sometimes it turns them off. Sometimes," she added, "it can allow a protein to translocate to different parts of a cell, or allow certain protein molecules to interact with other protein molecules. But at other times phosphorylation can prove very disruptive, as when it turns the tau protein from good to bad.

"It has been long speculated," the Harvard professor of pathology continued, "that a so-called aberrant phosphorylation, or hyperphosphorylation, of tau is one of the processes needed for the formation of neurofibrillary tangles - one of the hallmark lesions in AD." Thereby hangs the tale of a Jekyll and Hyde brain enzyme called cycline-dependent kinase 5 (Cdk5), and its partner protein, p35.

When Decent Molecules Go On A Phosphate High

Tsai narrated that story: "Cdk5 normally is a very important element in the development of the nervous system, and maintenance of adult brain function. It does that in part through its regulatory partner, the p35 protein. We, and a number of other laboratories, have shown that during embryonic development, Cdk5 is essential for the cytoarchitecture of the brain's cerebral cortex, cerebellum, hippocampus. And in adult neurons, normal Cdk5 actively participates in a number of key processes, such as neurotransmitter release and dopaminergic response. In short, Cdk5 is important for our day-to-day life, and brain function."

Cdk5 gets its wakeup calls from the up-close protein p35. Spawned by its gene on human chromosome 18, p35 comprises 307 amino acids, extending some 35 kiloDaltons long - whence its name.

"But the striking thing," Tsai went on, "is that apparently under those pathological conditions of hyperphosphorylation, a truncated form of p35 is produced, by its cleavage into two shorter products, one 10 kD in length, the other 25. And when this p25 protein is expressed," she warned, "things get very bad."

A paper in today's Nature, dated Dec. 9, 1999, of which Tsai is senior author, details her laboratory's findings, which update the clouded comprehension of the AD tangles. Its title: "Conversion of p35 to p25 deregulates Cdk5 activity and promotes neurodegeneration."

"In this article," Tsai said, "we show why it is bad. First of all, p35 has a very short half-life, and when it activates Ctk5, that activated enzyme triggers the degradation of p35, to make sure that Cdk5 is only transiently regulated. On the other hand, when p25 is produced, it's impossible to get rid of it. It's very stable, and it causes the Cdk5 to remain active indefinitely, and go to the site in the cell that normally Cdk5 doesn't visit. It's in the cytoplasm, just floating around, probably going to other organelles - but this work is still in progress. And p25 allows Cdk5 to phosphorylate substrates aberrantly, such as the tau protein.

"In this paper," Tsai said, "we show that this p25 protein, the truncated version of p35, accumulates to a very large extent in AD neurons containing neurofibrillary tangles. And that p25 is much more stable; it localizes to a completely different subcellular compartment. In addition, we also showed that the presence of p25 in primary neurons is very very bad - deleterious. It can cause the cytoskeleton of neurons to collapse, and it can cause profound cell death."

Human Brains Confirm Bad Effects

Tsai and her co-authors looked for these consequences in the brains of nine postmortem AD patients, compared with four controls, all in their 60s, 70s and 80s. They found AD neurons cluttered with tangles, particularly in the hippocampus, temporal frontal lobe and amygdala - the most vulnerable brain regions affected in the disease - but not in the brains of controls.

"I think the most obvious question," Tsai pointed out, "is whether this is going to be a good target for therapy or prevention of AD and perhaps other neurodegenerative diseases. I think it's an excellent target, especially the p25-Cdk5 connection. Already," she added, "there are companies trying to come out with specific Cdk5 inhibitors. I believe that probably will be helpful.

"But ideally, in the ideal world - though I don't know if it's possible," Tsai concluded, "is for people to come out with drugs that can selectively inhibit p25 activity, but without affecting p35."