24 October 2002

by Apoorva Mandavilli

The mood among Alzheimer's disease researchers last year was euphoric. Scientists were "most definitely, unquestionably" close to some real solutions, John Hardy, then director of the Center for Neuroscience at the Mayo Clinic in Florida, had claimed at the time. A year later, however, the tone is decidedly more somber.

A colleague of Hardy's, Michael Hutton, associate professor of neuroscience at the Mayo Clinic, was equally optimistic a year ago, but now admits that optimism was premature.

"There's a recognition that clinical trials and therapies aren't going to happen overnight," said Hutton. "It's going to be a longer process than we perhaps anticipated."

This time last year, drug company Elan was poised to begin Phase II trials of a much-publicized vaccine against the disease, while other companies touted promising inhibitors of enzymes critical in disease progression. The amyloid cascade hypothesis, which holds that deposition of amyloid-beta is the central event in the disease, seemed all but proven.

But in January, Elan, along with partner Wyeth-Ayerst Laboratories, suspended its trial after 15 participants developed central nervous system inflammation and an acute worsening of Alzheimer's symptoms. Although researchers have not abandoned the vaccine approach, there is much more discussion about precautions and procedures, and much less talk of a definitive cure.

"Everyone was disappointed that the vaccine was making people sick," said Hutton. Although the development is "not a huge setback," he added, "what we've seen over the last year is a more realistic attitude to the way therapy is going to work."

Among inhibitors to beta- and gamma-secretases, several of which were close to clinical trials last year, no single compound has emerged as leader. The amyloid debate - whether amyloid is a cause or by-product of the disease - is no closer to resolution, nor are other contentious questions in the field.

One prominent discussion has centered on the identity of gamma-secretase, which catalyzes the last step in amyloid-beta production. Based on solid biochemical and pharmacological evidence, many researchers have pointed to the protein presenilin, mutations in which are linked to familial Alzheimer's disease, as the elusive gamma-secretase.

In recent months, however, several different groups have identified at least three other proteins required for gamma-secretase activity. Inhibiting any of the four proteins - presenilin 1, nicastrin, pen-2, and aph-1 - eliminates enzyme action, researchers have found.

That immediately makes experiments much more complicated, says Gopal Thinakaran, assistant professor of neurobiology at the University of Chicago. For any cell biological experiment, scientists now have to introduce five proteins, including the amyloid precurosr protein (APP). "It's not an easy task," Thinakaran said. "The game is getting a little harder to play."

In a forthcoming issue of Neurobiology of Disease, Thinakaran and his colleagues further complicate the picture. Presenilin is not limited to catalytic activity and may fulfil multiple roles in different parts of the cell, they suggest.

Earlier experiments have shown that presenilin mutants (which lack both presenilin 1 and 2, encoded by the genes PS1 and PS2) fail to form the catalytic complex properly and show decreased gamma-secretase activity, although activity is not abolished.

Gamma-secretase normally cleaves the C-terminal fragment of APP, and the full-length peptide is not a substrate, notes Thinakaran. But, remarkably, he says, the PS1-PS2 double mutant affects maturation and trafficking of the full-length peptide.

"That tells us that presenilin is doing something to the full-length peptide, before it is cleaved by alpha- and beta-secretase," he said.

In the mutant, APP is hyper-glycosylated and greater numbers of the peptide leave the endoplasmic reticulum for the cell surface, suggesting that presenilin influences APP folding and maturation. "Presenilin clearly plays a co-chaperone type of role," Thinakaran said.

The researchers have not ruled out an enzymatic role for presenilin, however. "People always thought there was something they could not account for in terms of presenilin and APP," said Thinakaran. "It clearly plays multiple roles in the biology of APP."

The results "don't surprise me in the least," said Hutton. Whatever role presenilin may play, it's clear that gamma-secretase acts on substrates other than APP, he says, and one of those substrates may very well affect APP trafficking.

As scientists unravel such complex mechanisms in Alzheimer's disease, common threads have begun to emerge between AD and other neurodegenerative disorders including Parkinson's disease, Huntington's and amyotrophic lateral sclerosis (ALS).

For example, the E4 allele of apolipoprotein E (ApoE) is famously associated with the risk of Alzheimer's disease, with the severity of multiple sclerosis, and the age of onset of ALS.

In the largest study yet of ApoE and Parkinson's disease, neurologist Jeffrey Vance now reports significant associations between the E4 haplotype and disease occurrence in families with at least two affected members. Vance, who collaborated with researchers at GlaxoSmithKline, presented the results earlier this month at the American Society of Human Genetics (ASHG) meeting in Baltimore.

But what's the link between lipoproteins and nerve damage? "There's no hint of a mechanism," said Vance. "That's the Nobel Prize here." But as the list of implicated diseases grows, "we as scientists ought to be able to intersect those problems, and come up with a mechanism," he said.

One common pattern in neurodegenerative disorders is the presence of protein aggregates, such as amyloid plaques and Lewy bodies, which may be either a cause or by-product of the disease process.

The Lewy bodies in the brains of patients with Parkinson's disease contain alpha-synuclein, a protein of unknown function. Two known mutations in the protein are linked to familial forms of the disease. Now, neurologist Han-Xiang Deng, associate research professor of neurology at the University of Chicago, and his colleagues have created a transgenic mouse that they claim is the best mouse model for Parkinson's.

In the model, details of which are unpublished, mice express a dominant-negative mutant form of alpha-synuclein and develop Parkinson's-like symptoms, says Deng. At 300 days old, the mice have a full-blown neurological disorder characterized by bradykinesia, posture and gait problems.

Unlike other mouse models, Deng says, the mice also lost nearly half of all dopaminergic neurons. "This is the only [model] that shows typical phenotype and pathology," he said.

Interestingly, Deng's mice, and some other models of Parkinson's disease, do not have Lewy bodies, Deng notes. Based on those observations, "it's clear that Lewy body is not required for the disease," he said.

Although the mice do not have Lewy bodies, the model is an accurate reflection of the disease and is ideal for testing drugs such as p53 inhibitors, Deng adds. Using a different mouse model, researchers at the US National Institutes of Health are now reporting success with p53 inhibitors in preventing nerve damage.

If all goes well, human clinical trials could begin in two to three years, the NIH researchers say. But they caution that they need to make sure the inhibitors don't cause severe side effects.

After the suspension of the Elan trial, most researchers in the field are now more cautious about taking drugs to clinical trials, notes Hutton. Still, drugs aimed at amyloid-beta, the secretase enzymes, and other targets will eventually yield effective therapy, he says. "I still believe we're taking the right approach with all these drugs," said Hutton. "It's just a matter of time."

Send us your comments for publication.

Sign up for BioMedNet News weekly email alerts.