Immune molecules fine-tune brain circuits
by Parizad Bilimoria
Just as communication is critical to success in an organization or a relationship, communication between neurons is critical to all that our brains allow us to do—to read, speak, think, touch, feel, dream.
Synapses, the points where neurons communicate with each other, are building blocks of the brain’s circuits. There are literally trillions of them in the human brain. While some synapses live a long time, others appear transiently during brain development and are later pruned away. The loss of necessary synapses—or, conversely, the failure to prune ones that aren’t useful—is suspect in many brain diseases.
Beth Stevens, PhD, in the F.M. Kirby Center for Neurobiology at Children's Hospital Boston, has discovered some unexpected synapse eliminators: molecules of the innate immune system, traditionally our first-line defense against infection. Her new findings may help us better understand how neural circuits are fine-tuned—and how they fall apart or form incorrectly in neurological disorders as diverse as Alzheimer’s, glaucoma and epilepsy.
Old molecules, new field
In 2007, Stevens, then a postdoctoral fellow in the laboratory of Ben Barres, MD, PhD, Professor of Neurobiology at Stanford University, made a surprising discovery.
Work by Barres and others had revealed that glial cells, which greatly outnumber neurons in the brain—are just as important as neurons in shaping brain circuits. In particular, a star-shaped, multifunctional glial cell called the astrocyte releases factors that guide the development of synapses.
To learn more about these factors, Stevens and her labmates scanned gene activity in young neurons that were making synapses, to see which genes became more active when astrocytes were added to the culture dish. The activity of one gene – C1q – was dramatically increased. And interestingly, it was an immune gene that no one had expected to see in a normal brain. Previously it had only come up in the context of disease.
C1q is a classic innate immune molecule. It initiates the so-called complement cascade, a molecular pathway well known in immunology as a system for tagging unwanted cells and debris for clearance by other immune cells. In the brain, C1q appeared to be concentrated at developing synapses.
“Could complement be similarly tagging synapses the way it does a bacterial cell?” Stevens wondered.
Indeed, experiments in retinal neurons suggested that C1q and other members of the complement cascade mark certain synapses for elimination, calling out the weak links in brain circuits. Stevens and colleagues found extra synapses in the brains of mice that lacked C1q.
The team examined connections from the retina to the thalamus, a deep brain region that integrates sensory information. Some of these connections are known to be pruned during normal development of the visual system, but in the absence of C1q, many of these remained present and functional, as confirmed with electrophysiological recordings. The same thing happened in mice lacking a related immune molecule called C3, which follows C1q in the complement cascade.
Digging into disease
Realizing that innate immune molecules play an active role in development of the visual system, Stevens wondered about their role in brain disease. In both mouse models and the human brain, complement levels clearly rise in neurodegenerative diseases like glaucoma and Alzheimer’s. Synapse loss, one of the early events in neurodegenerative diseases, is thought to be important in their pathology.
Perhaps, Stevens thought, the high complement levels seen in diseased brains aren’t just vestiges of a general damage response. Perhaps the extra C1q actually instigates the destruction of stable synapses.
Stevens is now collaborating with laboratories that specialize in studying neurodegenerative diseases to explore this idea, in the visual system and beyond. “We’re picking diseases where synapse loss has already been described,” she explains, “and then trying to see if complement molecules are playing a role in that process.”
On the flip side, Stevens is exploring whether a failure of synapse elimination, and having less complement activity, could also underlie disease. “If this were global, meaning not just the visual system, one might hypothesize that having too many inputs might lead to hyperexcitability or even epilepsy,” she says.
Recently, in collaboration with Stevens and her former mentor Barres, David Prince, MD, of Stanford University, found that mice lacking C1q have increased synaptic connections in the neocortex, the part of the cerebral cortex that controls higher cognitive functions. These mice clearly suffer from epilepsy, suggesting that a failure in synapse pruning could indeed be a part of the disease.
Scavenging the synapse
Since immune molecules orchestrate the elimination of synapses during development, and since this process may be derailed or inappropriately activated by disease, it is important to figure out how exactly the elimination occurs. One major question is how C1q tags synapses for removal: How does one synapse get eliminated while its neighbors—often on the same neuron—stay intact?
Stevens hypothesizes that perhaps the complement proteins act as “punishment factors” which bind to unwanted, unstable synapses and call attention to them, while another set of molecules, the complement regulators, provide a “protective cloud” that shelters stable, necessary synapses in the vicinity.
A new line of investigation led by Dori Schafer, PhD, a postdoctoral fellow in Stevens’s laboratory, is testing the theory that the unstable synapses are pruned away by immune cells of the brain known as microglia.
Microglia are a prime suspect because they make C1q and other complement proteins, as well as the receptors for complement proteins. Also, they are activated during developmental synapse elimination. In other words, they carry the right equipment and are ready for duty at the right time to recognize synapses tagged for destruction.
If microglia do indeed prune unstable synapses during development, it wouldn’t be a stretch to think that they could also eliminate stable synapses during degenerative brain disease—synapses that should normally remain intact. While largely quiet in healthy adult brains, microglia and astrocytes quickly become reactive when the brain suffers an insult. “So the idea might be that when glia become reactive, this whole process of complement-dependent synapse elimination or synaptic remodeling may be recapitulated,” Stevens suggests.
The epilepsy in mice without C1q and, conversely, the high C1q levels in mice with neurodegenerative disease suggest that there are two sides to the immune equation in the brain. Perhaps complement proteins are slacking in disorders where there are too many unnecessary synapses and overzealous in diseases where synapses are being lost.
Interestingly, C1q and other complement proteins are not the only immune molecules discovered in the brain. Research teams are hotly pursuing others—such as the MHC1 proteins of the adaptive immune system, which allows the body to “remember” pathogens it has encountered—and looking at their role in synapse remodeling and refinement of brain circuits. “We talk a lot to see if there is a connection,” Stevens says.
Beyond improving our understanding of circuit refinement—one of the most fortuitous developments of the complement project, she adds—is the potential to shed light on a completely unanswered basic question. What do microglia do in the normal brain? The overwhelming majority of microglia studies to date have been conducted in the context of the injured brain, so it’s been hard to figure this out. By following the trail of the complement proteins during synapse development, her team may find some clues.