Our robust immune systems are equipped to conquer a host of infections, but sometimes the very cells sent to help cause catastrophic injury. This is a risk in viral meningitis, a relatively common infection of the cells lining the brain and spinal cord. In the process of tackling disease, immune cells sometimes damage blood vessels and cause them to leak, leading to convulsions and, ultimately, death. Until now, scientists believed that cytotoxic lymphocytes (CTLs), the killer cells that destroy infected cells, were causing this damage. But by obtaining a window onto the brain, Dorian McGavern at the Scripps Research Institute in La Jolla, California, Jiyun Kim of New York University School of Medicine and their colleagues show that CTLs merely recruit other immune cells that are responsible for wreaking the havoc.

To study the role of CTLs in viral meningitis, McGavern and his collaborators needed to find a way to see the CTLs in action in the living brain. Kim suggested performing a skull-thinning surgical technique in which a mouse's skull is painstakingly scraped until it is translucent enough for the brain to be visible through it, but still sturdy enough to provide protection for the underlying brain tissue. “It was ideal for viewing living brain tissue,” says Kim. “Usually researchers do a craniotomy, a hole in the skull.” But, she explains, that approach means exposing the brain to possible damage and inflammation.

The researchers then injected the mice with lymphocytic choriomeningitis virus, an infection that serves as a model of acute viral meningitis, and used two-photon microscopy to view CTLs in the animals' brains. The mice had been engineered so that these cells expressed a fluorescent tag. The authors were not expecting what they saw. “The vessels were breaking down,” says McGavern. “There was heavy damage and they were leaking. But when we looked at what the CTLs were doing, they weren't associated with that vascular damage at all.”

To determine what was causing the damage, the team first tested mutant mice deficient in various proteins typically produced by CTLs to kill their target cells. This essentially disarmed the CTLs, but to no avail. “Every time we knocked something out, we got the same result,” says McGavern. “The mice still died.”

The researchers discovered that, rather than causing vascular damage themselves, the CTLs were releasing chemoattractants that draw in myelomonocytic cells — a class of white blood cells that includes monocytes and neutrophils. Kim's live imaging showed that monocytes were tethered to the outside walls of brain blood vessels, apparently causing transient leakage throughout the brain. Neutrophils, conversely, lined up along the interior of the brain's blood vessels and burst through en masse, causing immediate injury (see page 191).

The finding that CTL-recruited myelomonocytes can produce the symptoms of meningitis has important implications. “The idea was, CTLs cause the damage and if you get rid of them, the damage goes away. The big surprise was that they weren't doing the physical damage; it was the cells they recruited. That was the twist,” McGavern says. These recruited cells may thus be a better target in the treatment of infections from West Nile virus to herpes simplex, both of which can cause meningitis, as well as multiple sclerosis and other autoimmune disorders — all conditions that involve damage to brain blood vessels by immune cells.