About five years ago, a team of Stanford University scientists set out to determine how the developing brain establishes its final set of synapses, connections through which cells of the nervous system communicate with one another and with nonneural cells. But when they tried to pinpoint the genes involved, something unexpected happened: they stumbled on C1q, a gene for a protein important in the body's immune system.

"We were like 'Wait a minute—this an immune system molecule. What's this doing in the brain?'" recalls lead researcher Ben Barres, a neurobiologist. "It stunned us."

Up until then, most scientists had believed that the healthy brain was "immune privileged" or free of immune cells. But Barres's team is one of several over the past decade that have not only discovered such molecules exist in the normal brain, but that they play a unique and essential role there. Experts say these findings provide a new window into the way the brain operates and why certain enigmatic disorders such as autism and Alzheimer's disease may develop, potentially paving the way for new therapies to treat them.

"The brain is invisible to certain kinds of immune surveillance," says Lisa Boulanger, a neurobiologist at the University of California, San Diego. "The whole body is under immune surveillance all the time. That surveillance is not as rapid or effective for infections in the brain. That was taken as evidence that the immune system was not present in the brain."

Researchers at Harvard University were the first to poke holes in this theory in a study published in 1998. A team led by neuroscientist Carla Shatz was running a routine procedure designed to identify genes regulated by neuronal firing when an unexpected one popped up that codes for major histocompatibility complex (MHC) class I molecules, which play a crucial role in helping the immune system recognize invading pathogens.

Further study revealed that these so-called immune proteins are actually present on the surface of certain nerve cells, but that they functioned differently in the brain than they did in the rest of the body; rather than scouting for germs, they influenced signals sent between neurons. "These molecules appear to control the way in which synaptic connections strengthen or weaken and how stable they are," Shatz says.

In particular, she says, they seem to act as a "brake" on synaptic plasticity or flexibility. "Mice that don’t have these molecules seem to be able to change their brain circuits with experience much more rapidly than normal mice," Shatz says. "We think these molecules are important for limiting connections," she adds, noting that if synapses are too malleable they can set off a host of problems, ranging from shaky brain circuits to seizures, which are caused by excessive neuronal firing.

Subsequent research by others has confirmed the importance of immune molecules in keeping the brain functioning smoothly. Humans are born with more synapses than necessary; as part of the normal development process, weak and unnecessary connections are gradually eliminated during childhood. Barres and his colleagues found that production of C1q—the immune protein they unexpectedly found in the brain—peaks at the same time that synapses are pruned. What's more, animals that lack the protein have extraneous connections in their brains even as adults.

Upon further examination, Barres, in conjunction with Simon John—a neuroscientist at Jackson Laboratory in Bar Harbor, Me.—discovered that C1q appears to be involved in synapse loss linked to glaucoma. Mice with this eye disease, which damages the optic nerve and causes vision loss, have higher levels of the immune molecule, which accumulates at retinal synapses before the neurons die.

The scientists believe that immune proteins in the brain may be so significant that disruption during development might also contribute to such conditions as autism and schizophrenia. Research has shown, for instance, that if a woman is exposed to a virus while pregnant, it could increase the odds that her child will develop one of these disorders. Boulanger speculates that immune molecules in the brain may be the link.

When a pregnant woman is exposed to a virus, it activates her immune response and influences the level of certain immune proteins in the body of her fetus. But in the fetal brain, some of the immune molecules are otherwise occupied, helping with synapse formation and remodeling. "These molecules have a kind of 'night job' in the brain and it's completely different than their immune jobs," Boulanger says. During fetal development, "the molecules are busy building the brain—you don't want to disrupt them." If the molecules are pulled into an immune function in the fetus, she says, it could disrupt brain development and lead to neurodevelopmental disorders.

Boulanger recently found that decreasing the level of the MHC class I molecules in the bodies and brains of mice was sufficient to cause biochemical and behavioral symptoms of autism and schizophrenia in the mice.  She is now investigating whether this is also the case in humans by studying samples of bodily fluids of autistic and schizophrenic patients (as well as brain tissue extracted during autopsies) for evidence of abnormal MHC class I levels.

If the connection holds up, it could provide important new insights into the cause of two of the most mysterious neurological disorders. "It could also point the way to a completely weird and unexpected strategy for drug development," Boulanger says. "Maybe modifying immune signaling would help some of these patients."

But it is not just the function of immunological molecules in the developing brain that the new findings might help illuminate—they could also help explain the molecular processes involved in neurodegeneration, the researchers say. Synapse loss is typical of the early stages of Alzheimer's and similar disorders. Barres speculates that the same immune molecules that help eliminate unneeded synapses in the young brain may mistakenly destroy necessary synapses later in life, causing the symptoms of these neurodegenerative diseases.

"Is this pathway of the normal developing brain somehow reactivated in the adult brain in the case of neuro-injury or neurodegenerative disease?" he says.

The bottom line: finding connections between immune molecules and other kinds of brain degeneration and damage could create a whole new category of treatment targets.

"The door is really wide open to a new way of thinking about brain degeneration," Shatz says, adding that the intersection of immunology and neurology warrants a lot more research. "It's [the] early days. We're all excited about the convergence of the two fields."