Apparently APP

Axons lose surface APP in the absence of trophic factors.
Images Courtesy of A. Nikolaev and M. Tessier-Lavigne

Ten years ago, Marc Tessier-Lavigne, a neurobiologist then at the University of California, San Francisco, saw an image he's never forgotten. His postdoc Zhigang He (now at Harvard) showed him a picture of stained mouse embryos indicating that the beta-amyloid precursor protein (APP), a known "bad actor" in Alzheimer's disease, was highly enriched in neurons and axons during development. Although Tessier-Lavigne never published the finding, the snapshot of the red-tinged embryos "got seared in my brain at the time," he says, telling him that APP might be involved fundamentally in axonal growth or guidance. That memory eventually—and fortuitously—ended up pointing his research towards a potential treatment for the disease.

A couple years back, Tessier-Lavigne and his postdoc Anatoly Nikolaev were studying neuronal expression levels of some 30-odd immunity-related receptors, when they noticed that one—the little-studied "death receptor" DR6—was widely expressed right at the time when neurons are born. They knew from previous work that early neurons would commit suicide if not bathed with supporting "trophic" factors, and they wondered if this death receptor—so called because it contains a cytoplasmic death domain—might be involved in neuronal death.

Tessier-Lavigne's unforgettable image: A section through the spinal cord (a) and dorsal root ganglia (b) of a mouse embryo, stained to reveal the sites of expression of APP.

They used RNA interference to knock down DR6 expression in mouse and rat embryos and discovered that spinal neurons no longer degenerated during development. Anti-DR6 antibodies produced the same effect; so did genetic deletions. "That was the first indication that we were on to something interesting," and that this receptor might be a key player in the elusive process of neurodegeneration, says Tessier-Lavigne, now executive vice president of research drug discovery at Genentech in South San Francisco.

Then, they thought: If DR6 is a receptor, surely it has a ligand. A series of experiments proved this was the case: A soluble version of DR6 mopped up an unknown ligand, which they then showed was present on the neuronal surface. And when they took away the neuron's trophic factors, the ligand was released in a highly active form that doubled back on DR6 to kill the neurons (Nature, 457:981–87, 2009).

The big question then was: What's the elusive ligand? "We thought, 'Wait a minute; what about APP?'" says Tessier-Lavigne. In the decade-old images, he had seen that APP was expressed on the surface of axons. Plus, he knew that APP's outer domain could be cleaved and released, and that the protein was tied to neurodegeneration through its links with Alzheimer's disease. "Those three things made us think that maybe APP was the ligand that we were looking for," he says.

Indeed it was. "We fell off our chairs repeatedly when one experiment after another led us to believe that the ligand was our old friend APP," Tessier-Lavigne says. "The potential therapeutic applications were immediately obvious. We knew we were onto something really important."

After neurons have elaborated their branches, they undergo a process of axon pruning during which APP is broken down to form two main components: the N-terminal fragment, called N-APP, which binds DR6 and triggers neuron death; and amyloid-beta peptide, the main driver of Alzheimer's-related plaque formation. Considering the link between neurodegeneration and Alzheimer's, Tessier-Lavigne proposed that N-APP may contribute to the initiation or progression of the disease, either alone or in collaboration with amyloid-beta, and that both components might unleash a double whammy that wipes out nerve cells in the debilitating disease.

"We know that Alzheimer's disease hijacks APP," he says. "If it hijacks the molecule, wouldn't it make sense to hijack the mechanism?" If so, then N-APP should make a good new drug target to combat Alzheimer's, he adds.

Marc Freeman, a neurobiologist at the University of Massachusetts Medical School in Worcester, flags another compelling aspect of the study: Unlike neuronal cell body degeneration, which requires the apoptosis-related enzyme caspase 3, Tessier-Lavigne showed that the DR6-mediated pathway requires a different enzyme, caspase 6. "The concept in the field has been that caspase activation is not important in axonal degradation," says Freeman. "But Marc [Tessier-Lavigne]'s work shows very clearly that one way you can destroy an axon is through caspase activation."

The calamitous caspase also contributes to the progression of Huntington's disease, and this overlap is "starting to triangulate the biological mechanisms" that are involved in seemingly different neurodegenerative diseases, says Don Nicholson, franchise worldwide basic research head at Merck Research Laboratories in Rahway, NJ.