Plaques and tangles pockmark the brains of people with Alzheimer’s disease. The extracellular protein amyloid-β makes plaques, and the intracellular protein tau makes tangles, but how exactly these might kill neurons is unclear. Work presented at the annual meeting of the American Society for Cell Biology in San Francisco, California, this week starts to connect some of these dots.

George Bloom, of the University of Virginia in Charlottesville, and his colleagues began by following up on work that neurons exposed to amyloid-β die not from direct poisoning, but because amyloid-β prompts inappropriate cell behaviour. They re-enter the cell cycle but never divide, and die instead.

“The framework of the process has now been defined,” he says. “We think we’ve stumbled upon one of the seminal events in the transition of healthy neurons into Alzheimer neurons.”

The work identifies several potential very early biomarkers of Alzheimer’s disease and suggests new ideas to treat it.

Within 24 hours of exposing neurons to amyloid-β, Bloom and his students could see that the neurons had begun to duplicate DNA — an early preparation for division. When they repeated the experiment with neurons that lacked tau, however, the neurons did not respond this way.

Guessing that a cell-signaling cascade could explain these observations, Bloom’s then-student Matt Seward began listing potential protein mediators. He identified enzymes called kinases that had been implicated in Alzheimer’s disease, cell-cycle regulation or tau modification.

This revealed three kinases (Fyn, CaMKII and PKA) that each had to be activated for the neurons to re-enter the cell cycle. Each of these kinases modifies tau at a particular site. If any of those sites were mutated, amyloid-β no longer prompted neurons to re-enter the cell cycle.

To figure out whether this process might happen in actual brains, Bloom collaborated with Eric Roberson at the University of Alabama, Birmingham, who has bred mice with an amyloid-β mutation that show learning and memory problems reminiscent of Alzhiemer’s disease. He has also crossed this mutation into mice that cannot express tau and found that although the mice still develop amyloid plaques, they do not display learning or memory problems.

When Bloom examined sections of brains from these mice, he found that neurons in brains expressing tau also expressed proteins made during cell division. In mice that lacked tau, those proteins were absent.

Several papers have been published suggesting that “amyloid is the trigger and tau is the bullet”, says Bloom, but the connection to the cell cycle and how amyloid-β might affect tau was less clear. There are still many steps to be sussed out. Nonetheless, he says, “this is the first time anyone has started to work out the biochemistry”.