Microglia are the nervous system's exclusive immune cells. In a resting state, they resemble long-legged spiders, but when confronted with pathogens or injury, they retract their appendages and balloon into round blobs that engulf and remove the pathogens and other cellular debris. They also eliminate damaged synapses. “The idea that [microglia] can clean up brain debris has been well established in studies of brain disease,” Beth Stevens (Boston Children's Hospital) told New Scientist. “But now, even without damage, we've found them to respond to subtle changes in synaptic function.” Stevens led a study in mice that explored more closely the activity of microglia in healthy brain tissue. The results, recently published in (Neuron 74, 691–702; 2012), contradict the notion that microglia are primarily passive immune guardians in the brain.

Stevens and her colleagues designed assays for imaging the neural circuitry of the visual system. They labeled the neurons projecting from each eye into the brain with different colors. The visual system undergoes active change during development, and, in keeping with the idea that microglia remove unneeded synapses, Stevens' group observed labeled pieces of synaptic material inside activated microglia in the brains of the mice. Next, they chemically modified the amount of neural activity coming from each eye, altering the strength of the synapses associated with each eye. Microglia contained more synaptic fragments from the more active eye than from the less active eye, leading the group to conclude that microglia can detect and respond to a synapse's level of activity, selectively destroying less active circuits, even in healthy brains.

But the molecular signal that marked a synapse for destruction by microglia remained a mystery. Stevens' group suspected that the complement system might be involved. Previous work had suggested that the complement system could 'tag' weak synapses. When Stevens' team repeated their experiments in mice that lacked the complement receptor C3, they found that the microglia engulfed significantly fewer synapses. “We think C3 is an 'eat me' signal,” said Stevens, according to New Scientist.

Although the findings provide a better understanding of how synapses are maintained, scientists are still uncertain whether microglia play an active part in their maintenance. Microglia seem to be closely associated with the process but have never been 'caught in the act' of synaptic pruning.

Nevertheless, the results raise the possibility that manipulating microglia could be a new avenue for treating certain brain disorders, such as those associated with abnormal brain development or early synapse loss.