Key Alzheimer's findings questioned

Conflicting results cloud link to prion protein.

Plaques can destroy nerve cells and cause the brain to shrink in Alzheimer's disease (left) — but how? Credit: PASIEKA/SPL

When Stephen Strittmatter discovered an unexpected link between key proteins in two devastating brain maladies¹ — Alzheimer's disease and Creutzfeldt–Jakob disease (CJD) — researchers in the field agreed that he was on to something big. But a year on, conflicting results, including findings published in Nature this month², have clouded that rosy picture and highlight the challenge faced by researchers seeking a way to arrest Alzheimer's disease progression.

Strittmatter, a neurologist at Yale University in New Haven, and his team were trying to understand how the protein fragment amyloid-β causes damage when it forms plaques in the brains of people with Alzheimer's disease. They found that cellular prion protein, whose abnormal, infectious form is infamously associated with CJD, can bind readily to amyloid-β, suggesting that it may act as a receptor that kicks off a chain of events leading to neuronal destruction¹.

"It really was a spectacular finding," says Bart De Strooper, who studies neurodegenerative diseases at the Catholic University of Leuven in Belgium.

So far, therapies that directly target amyloid-β have not performed well (see 'Clinical trial disappoints'). The prospect of targeting prions instead was a cause for excitement. "It could make a fast jump to a real therapy, if it were a real target," says Dominic Walsh, who studies Alzheimer's disease at University College Dublin.

But the disparate experimental methods used by researchers often lead to contradictory conclusions. "There's just a history of this in the field," says Roberto Malinow, an electrophysiologist at the University of California, San Diego. "There are these spectacular results and they have a half-life of six months to a couple of years."

Strittmatter and his team speculated that the prion protein, which is present on the surfaces of neurons, might activate a cell-damaging pathway when bound and activated by amyloid-β. To test that idea, they injected amyloid-β into mice and measured the electrical changes that take place in the brain during memory retrieval. The electrical signatures were unaffected in mice that lacked the prion protein — suggesting that the prion protein mediates damage caused by amyloid-β.

Yet when others in the field rushed to build on the work, the results differed. Gianluigi Forloni, an Alzheimer's researcher at the Mario Negri Institute for Pharmacological Research in Milan, Italy, and his team also injected amyloid-β into mice lacking the prion protein, but found that the mice suffered memory deficits even in the absence of the prion protein³.

In Switzerland, Adriano Aguzzi, a prion-protein researcher at University Hospital Zürich, and his group reached the same conclusion⁴ using mice that were genetically engineered to produce copious amounts of amyloid-β. And Malinow and his colleagues closely replicated the brain measurements of Strittmatter and his team, but failed to replicate their findings².

Strittmatter cautions that differences in the groups' animal models, memory tests and methods of preparing and administering amyloid-β may explain the conflicting results. Forloni takes a more negative view. "If it depends so much on the original experimental conditions, then the findings become no longer necessarily relevant."

In more recent work⁵, Strittmatter and his colleagues report that a lack of prion protein staved off memory loss due to amyloid-β accumulation in mice that are genetically different from those that Aguzzi used. And in July, a team lead by Michael Rowan, a neuropharmacologist at Trinity College Dublin, reported at the Federation of European Neuroscience Societies meeting in Amsterdam that antibodies that block access to the prion protein could reduce the effect of amyloid-β on electrical changes in the brain associated with memory, backing up the findings of Strittmatter and his team.

Conflicting results are common in Alzheimer's disease research, says Sangram Sisodia, a neurobiologist at the University of Chicago in Illinois. Animal models are particularly contentious, he notes, because there is no single model that is universally perceived to be the best system in which to study the disease. And because each lab has invested its time, effort and reputation into its particular model system, this is unlikely to change, he adds. A consortium organized by the Cure Alzheimer's Fund, based in Boston, Massachusetts, is trying to unite members of the competitive and often contentious field. "But in terms of standardizing experimental approaches, that is a long ways off," says Sisodia. "I just don't see that happening."

Box 1: Clinical trial disappoints

The next great hope in the fight against Alzheimer’s disease met an untimely end last week, when safety concerns forced drug maker Eli Lilly, headquartered in Indianapolis, Indiana, to halt two large clinical trials.

The drug semagacestat was designed to reduce the production of amyloid-β plaques in the brain — a defining feature of Alzheimer’s disease. But preliminary data from more than 2,600 patients showed that the drug increased the risk of skin cancer, and what’s more, it seemed to worsen memory rather than protect it.

The results came as no surprise to some observers. Lilly tested the drug in patients with mild-to-moderate Alzheimer’s disease, when damage in the brain may be too advanced for treatment.

Semagacestat blocks the enzyme γ-secretase, which helps to produce amyloid-β. But γ-secretase also processes several other important proteins, including epidermal growth factor receptor — the loss of which causes skin cancer in animal models.

Despite the failure, amyloid-β will probably remain a key target for drug development, says Paul Aisen, an Alzheimer’s researcher at the University of California, San Diego. “The Lilly trial was not good news,” he says. “But therapies that target amyloid are still at the top of the pile."