By Ken Garber

Science, April 25, 2014

In 2004, veteran structural biologists Gregory Petsko and Dagmar Ringe were looking for a way to make an impact on Alzheimer’s disease. Frustration ensued. “We couldn’t see a fresh approach,” Petsko says.

Then, at a meeting in Colorado, they heard Scott Small, a young Columbia University neurologist, describe his research probing the brains of people who had died of the neurodegenerative disorder. “Along comes Scott with a totally new pathway,” involving an obscure molecular complex called a retromer, Petsko says. “This really got our attention.”

Their lunch that day with Small set the agenda for a decadelong collaboration that this week unveils, in Nature Chemical Biology, a potential new drug lead for Alzheimer’s. By stabilizing the components of retromers, which act like recycling bins in cells, the compound could slow or stop the implacable spread of a protein fragment called beta amyloid, a likely culprit in neuron death in Alzheimer’s. Although preliminary— so far the published work has been done only in cells—the new results have nonetheless impressed some veterans of the Alzheimer’s field.

What Small found a decade ago was that in the hippocampus, a key brain region for memory processing, one area typically ravaged by Alzheimer’s disease contained less of a key retromer component than a resistant area. The retromer is a linked set of proteins that encloses cargoes of cell surface proteins, including beta amyloid’s precursor, helping to recycle them from the cell interior to the outer membrane
(see diagram).

Small’s discovery that retromers are lacking in the precise area of the brain fi rst affected by Alzheimer’s disease offered a possible solution to one of the condition’s biggest puzzles. In the roughly 99% of Alzheimer’s cases that lack a known genetic cause, no one knows why β amyloid builds up in the brain. For the fragment to exist, certain enzymes must cleave its parent, amyloid precursor protein (APP). This is thought to happen in an organelle called the endosome. Based on his brain studies, Small hypothesized that a lack of retromer activity causes more APP (and other proteins) to linger in endosomes, giving enzymes greater opportunity to generate beta amyloid. (Enlarged endosomes are often seen in Alzheimer’s disease.)

“You have APP residing for too long in the endosome,” Small says. “And so—bam—that’s where the enzymes start cleaving.”

In 2004, however, it wasn’t clear if the retromer defect was a cause or consequence of the brain disorder. Evidence for causality has accumulated over the last decade. Knocking out one copy of a retromer gene in mice leads to higher beta amyloid and to memory deficits, and variants in several retromer-related genes are associated with Alzheimer’s risk. “Some of the story remains speculative,”says neurologist Samuel Gandy of the Mount Sinai School of Medicine in New York City. “Importantly, nothing has come along that disproves the model.”

While Small was building the case for retromer dysfunction in Alzheimer’s, Petsko, Ringe, and graduate student Vincent Mecozzi, all at Brandeis University in Waltham, Massachusetts,set out to design a “chaperone”compound that would stabilize the complex retromer structure and thereby boost overall retromer numbers. Using structural biology tools, they identifi ed a weak point in the linked retromer proteins where a single compound might bind them more tightly together, thus stabilizing the complex and slowing typical degradation in a cell. After several years of work, and somewhat to their own surprise, they found such a compound.

In cultured hippocampal neurons, the small molecule boosts the amount of retromer proteins by 50% or more, and reduces levels of human beta amyloid by 39% in neurons from mice expressing a mutant APP seen in Alzheimer’s disease. “The new proretromer drug is brilliant. … If you had asked me whether I thought the retromer could be tweaked specifi cally, I would have guessed no,” Gandy says. “Time (and trials) will tell whether it has true promise as a therapeutic.”

Indeed, indiscriminate recycling of molecules in cells could cause side effects—APP isn’t the only cargo retromers carry. The goal of treatment, Small says, is to tweak the system to bring retromer activity back to normal levels, but he admits that such a drug could end up increasing cargo “flow” in normal cells, which could be harmful—no one knows. “These are all incredibly important questions that we’re actively involved with,” Small says.

The compound’s success also depends on endosomes being the site of most APP processing. Most studies confirm this, but a few locate beta amyloid production in the Golgi apparatus, an organelle that packages proteins for delivery elsewhere. If beta amyloid is mainly produced there, a pro-retromer drug could exacerbate Alzheimer’s disease by delivering more APP.

A retromer-boosting drug may also work in Parkinson’s disease. In 2011, two papers revealed that rare mutations in a retromer gene cause that neurodegenerative disorder in families. “That was a big surprise to all of us,” says Petsko, who’s moving to Weill Cornell Medical College in New York City, where he’ll try to identify the critical retromer cargo in Parkinson’s neurons.

The group is now raising money for further drug discovery and development— they’ve started to give their compound to mice. The retromer team knows the long odds of developing a safe and effective drug for Alzheimer’s disease, let alone for two brain conditions. “Most attempts at therapy fail,” Petsko says. “I can live with that. I couldn’t live with not trying.”