Overconsumption of synapses by immune cells may drive early-stage Alzheimer’s disease.

Over-pruning synapses may drive early-stage Alzheimer’s disease.

Eraxion/iStockphoto

Alzheimer’s may be caused by haywire immune system eating brain connections

More than 99% of clinical trials for Alzheimer’s drugs have failed, leading many to wonder whether pharmaceutical companies have gone after the wrong targets. Now, research in mice points to a potential new target: a developmental process gone awry, which causes some immune cells to feast on the connections between neurons.

“It is beautiful new work,” which “brings into light what’s happening in the early stage of the disease,” says Jonathan Kipnis, a neuroscientist at the University of Virginia School of Medicine in Charlottesville.

Most new Alzheimer’s drugs aim to eliminate β amyloid, a protein that forms telltale sticky plaques around neurons in people with the disease. Those with Alzheimer’s tend to have more of these deposits in their brains than do healthy people, yet more plaques don’t always mean more severe symptoms such as memory loss or poor attention, says Beth Stevens of Boston Children’s Hospital, who led the new work.

What does track well with the cognitive decline seen in Alzheimer’s disease—at least in mice that carry genes that confer high risk for the condition in people—is a marked loss of synapses, particularly in brain regions key to memory, Stevens says. These junctions between nerve cells are where neurotransmitters are released to spark the brain’s electrical activity.

Stevens has spent much of her career studying a normal immune mechanism that prunes weak or unnecessary synapses as the brain matures from the womb through adolescence, allowing more important connections to become stronger. In this process, a protein called C1q sets off a series of chemical reactions that ultimately mark a synapse for destruction. After a synapse has been “tagged,” immune cells called microglia—the brain’s trash disposal service—know to “eat” it, Stevens says. When this system goes awry during the brain’s development, whether in the womb or later during childhood and into the teen years, it may lead to psychiatric disorders such as schizophrenia, she says.

Stevens hypothesized that the same mechanism goes awry in early Alzheimer’s disease, leading to the destruction of good synapses and ultimately to cognitive impairment. Using two Alzheimer’s mouse models—each of which produces excess amounts of the β amyloid protein, and develops memory and learning impairments as they age—she and her team found that both strains had elevated levels of C1q in brain tissue. When they used an antibody to block C1q from setting off the microglial feast, however, synapse loss did not occur, the team reports today in Science.

To Stevens, that suggests that the normal mechanism for pruning synapses during development somehow gets turned back on again in the adult brain in Alzheimer’s, with dangerous consequences. “Instead of nicely whittling away [at synapses], microglia are eating when they’re not supposed to,” she says.

The group is now tracking these mice to see whether a drug that blocks C1q slows their cognitive decline. To determine whether elevated β amyloid can cause the C1q system to go haywire, Stevens and colleagues also injected a form of the protein which is known to generate plaques into the brains of normal mice and so-called knockouts that could not produce C1q because of a genetic mutation. Although normal mice exposed to the protein lost many synapses, knockouts were largely unaffected, Stevens says. In addition, microglia only went after synapses when β amyloid was present, suggesting that the combination of protein and C1q is what destroys synapses, rather than either element alone, she says, adding that other triggers, such as inflammatory molecules called cytokines, might also set the system off.

The findings contradict earlier theories which held that increased microglia and C1q activity were merely part of an inflammatory reaction to β amyloid plaques. Instead, microglia seem to start gorging on synapses long before plaques form, Stevens says. She and several co-authors are shareholders in Annexon Biosciences, a biotechnology company that will soon start testing the safety of a human form of the antibody the team used to block C1q, known as ANX-005, in people.

Such a central role for microglia in Alzheimer’s disease is “still on the controversial side,” says Edward Ruthazer, a neuroscientist at the Montreal Neurological Institute and Hospital in Canada. One “really compelling” sign that the mechanism is important in people would be if high levels of C1q in cerebrospinal fluid early on predicted developing full-blown Alzheimer’s later in life, he says. Still, he says, “it’s difficult to argue with the strength of the study’s evidence.”